http://www.cambridge.org/9780521883559

This page intentionally left blank

Personal Epistemology in the Classroom

Personal epistemology is the study of beliefs associated with know­ledge and knowing. A large body of theory and research in personal epistemology has been dedicated to college students, but rarely have the epistemic beliefs of children, adolescents, and their teachers been thoroughly examined. This book incorporates both theoretical and empirical work pertaining to personal epistemology as it specifically relates to learning and instruction. Bringing together leading research on preschool through high school students’ personal epistemology, it re­examines existing conceptual frameworks, introduces new models, provides an empirical foundation for learning and instruction, and considers broader educational implications. In addition, the contribu­tors stress how personal epistemology issues in the classroom need to be more carefully investigated and understood.

lisa d. bendixen is an Associate Professor in the Department of Educational Psychology at the University of Nevada, Las Vegas.

florian c. feucht is an Assistant Professor in Educational Psych­ology in the Department of Educational Foundations and Leadership in the Judith Herb College of Education, University of Toledo, Ohio.

Personal Epistemology in the Classroom: Theory, Research, and Implications for Practice

Lisa D. Bendixen and Florian C. Feucht

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore,

São Paulo, Delhi, Dubai, Tokyo

Cambridge University Press

The Edinburgh Building, Cambridge CB2 8RU, UK

First published in print format

ISBN-13 978-0-521-88355-9

ISBN-13 978-0-511-69148-5

© Cambridge University Press 2010

2010

Information on this title: www.cambridge.org/9780521883559

This publication is in copyright. Subject to statutory exception and to the

provision of relevant collective licensing agreements, no reproduction of any part

may take place without the written permission of Cambridge University Press.

Cambridge University Press has no responsibility for the persistence or accuracy

of urls for external or third-party internet websites referred to in this publication,

and does not guarantee that any content on such websites is, or will remain,

accurate or appropriate.

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org

eBook (NetLibrary)

Hardback

We would like to dedicate this book to our grandmothers, Maribel S. Page (sitting at her one­room school house desk) and Elisabeth Feucht (with a goody basket on her first day of elementary school), for

their loving support of us, especially in our educational pursuits.

vii

Contents

List of figures page xList of tables xiiContributors xiv

Part I Introduction 1

1 Personal epistemology in the classroom: a welcome and guide for the reader 3f l or i a n c. f e uc h t a n d l i s a d. b e n di x e n

Part II Frameworks and conceptual issues 29

2 Manifestations of an epistemological belief system in preschool to grade twelve classrooms 31m a r l e n e sc hom m e r-a i k i n s, m a ry bi r d, a n d l i n da b a k k e n

3 Epistemic climate in elementary classrooms 55f l or i a n c. f e uc h t

4 The integrative model of personal epistemology development: theoretical underpinnings and implications for education 94de a n n a c. ru l e a n d l i s a d. b e n di x e n

5 An epistemic framework for scientific reasoning in informal contexts 124fa ng -y i ng ya ng a n d c h i n- c h u ng t s a i

6 Who knows what and who can we believe? Epistemological beliefs are beliefs about knowledge (mostly) to be attained from others 163r a i n e r brom m e , d oro t h e k i e n h u e s, a n d t or s t e n p or sc h

Contentsviii

Part III Students’ personal epistemology, its development, and its relation to learning 195

7 Stalking young persons’ changing beliefs about belief 197m ic h a e l j. c h a n dl e r a n d t r av i s p rou l x

8 Epistemological development in very young knowers 220l e a h k. w i l de ng e r, b a r b a r a k. hof e r, a n d j e a n e . bu r r

9 Beliefs about knowledge and revision of knowledge: on the importance of epistemic beliefs for intentional conceptual change in elementary and middle school students 258l uc i a m a s on

10 The reflexive relation between students’ mathematics­related beliefs and the mathematics classroom culture 292e r i k de c or t e , p e t e r op ’t e y n de , f i e n de pa e p e , a n d l i e v e n v e r sc h a f f e l

11 Examining the influence of epistemic beliefs and goal orientations on the academic performance of adolescent students enrolled in high­poverty, high­minority schools 328p. k a r e n m u r p h y, m ic h e l l e m. bu e h l , j i l l a . z e ru t h, m a e g h a n n. e dwa r d s, j oyc e f. l ong, a n d s h i n ic h i monoi

12 Using cognitive interviewing to explore elementary and secondary school students’ epistemic and ontological cognition 368j e f f r e y a. gr e e n e , j u di t h t or n e y-p u r ta, ro g e r a z e v e d o, a n d j a n e rob e r t s on

Part IV Teachers’ personal epistemology and its impact on classroom teaching 407

13 Epistemological resources and framing: a cognitive framework for helping teachers interpret and respond to their students’ epistemologies 409a n dr e w e l by a n d dav i d h a m m e r

14 The effects of teachers’ beliefs on elementary students’ beliefs, motivation, and achievement in mathematics 435k r i s ta r . m u i s a n d m ic h a e l j. f oy

15 Teachers’ articulation of beliefs about teaching knowledge: conceptualizing a belief framework 470h e l e n ro s e f i v e s a n d m ic h e l l e m. bu e h l

ix

16 Beyond epistemology: assessing teachers’ epistemological and ontological worldviews 516l or i ol a f s on a n d gr e g ory sc h r aw

Part V Conclusion 553

17 Personal epistemology in the classroom: what does research and theory tell us and where do we need to go next? 555l i s a d. b e n di x e n a n d f l or i a n c. f e uc h t

Index 587

Contents

x

Figures

3.1 The educational model of personal epistemology page 59 4.1 The integrative model for personal epistemology

development 98 4.2 Key classroom elements for epistemic advancement 115 5.1 Theoretical framework for the development of scientific

reasoning in informal contexts and personal epistemology 146

6.1 Means of external sources ratings subdivided into curricular topics 180

6.2 Mean source preferences for math and Sachkunde lesson context 181

8.1 Proposed timeline for epistemological development across childhood 227

9.1 Mean scores for conceptual change about biological evolution by epistemic beliefs 271

9.2 Mean scores for metaconceptual awareness by epistemic beliefs and type of text 272

9.3 Overall mean scores for conceptual change about light showing the interaction between epistemic beliefs, topic interest, and text type 274

9.4 Mean scores for conceptual change about magnets by epistemic beliefs and mastery goal orientation 275

11.1 Hypothesized model for the relationship among students’ epistemic beliefs, achievement goals, and academic achievement 354

11.2 Eighth­grade model with epistemic beliefs, two­factor goals, and students’ GPA 355

11.3 Ninth­grade model with epistemic beliefs, three­factor goals, and students’ GPA 356

12.1 Relative location of academic domains on secondary students’ belief continua 393

14.1 Hypothesized model 445

xi

14.2 Final model 45116.1 The four­quadrant scale 52516.2 Visual results 53017.1 Six themes related to personal epistemology in the

classroom using a systems perspective 556

List of Figures

xii

Tables

1.1 Overview of the book page 14 3.1 Overview of models describing characteristics

of epistemic climate 61 5.1 Descriptions for the “view of knowledge” in the

Perry scheme 128 5.2 Percentage distributions of epistemological perspectives

regarding the nature of knowledge 141 5.3 Percentage distributions of epistemological perspectives

regarding the process of knowing 141 5.4 Performances of coordinating theory and evidence 142 5.5 Performance of reflective thinking 143 5.6 Effects of personal epistemology on scientific reasoning 144 6.1 Study design 179 7.1 Two imagined epistemic dimensions 200 8.1 Prior domain­dependency research on personal

epistemology with children 230 9.1 Summary of regression analysis for conceptual change

learning from text in fifth­graders 27811.1 Items assessing each of the factors for eighth­ and

ninth­grade participants 34011.2 Correlations among goal orientations for eighth­grade

students 34511.3 Pattern coefficients for the eighth­ and ninth­grade

goals factors 34711.4 Correlations among epistemic beliefs, goal orientations,

and GPA for eighth­grade students 34911.5 Correlations among goal orientations for

ninth­grade students 34911.6 Correlations among epistemic beliefs, goal orientations,

and GPA for ninth­grade students 35112.1 Model of epistemic and ontological cognition 376

xiiiList of tables

12.2 Participant demographic information 38212.3 EBI wording and EOCQ wording 38312.4 Examples of prewritten semi­structured interview probes 38514.1 Means, standard deviations, and alpha coefficients 45015.1 Framework of beliefs about teaching knowledge 48016.1 Key terms, definitions, and sources 519

xiv

Contributors

ro g e r a z e v e d o, University of Memphis

l i n da b a k k e n, Wichita State University

l i s a d. b e n di x e n, University of Nevada, Las Vegas

m a ry bi r d, Wichita State University

r a i n e r brom m e , University of Muenster

m ic h e l l e m. bu e h l , George Mason University

j e a n e . bu r r, Hamilton College

m ic h a e l j. c h a n dl e r, University of British Columbia

e r i k de c or t e , University of Leuven

f i e n de pa e p e , University of Leuven

m a e g h a n n. e dwa r d s, University of Oklahoma

a n dr e w e l by, University of Maryland

f l or i a n c. f e uc h t, University of Toledo

h e l e n ro s e f i v e s, Montclair State University

m ic h a e l j. f oy, McGill University

j e f f r e y a. gr e e n e , University of North Carolina at Chapel Hill

dav i d h a m m e r, University of Maryland

b a r b a r a k. hof e r, Middlebury College

d oro t h e k i e n h u e s, University of Muenster

j oyc e f. l ong, University of Notre Dame

l uc i a m a s on, University of Padua

xvContributors

s h i n ic h i monoi, Ohio State University

k r i s ta r . m u i s , McGill University

p. k a r e n m u r p h y, The Pennsylvania State University

l or i ol a f s on, University of Nevada, Las Vegas

p e t e r op ’t e y n de , University of Leuven

t or s t e n p or sc h, University of Muenster

t r av i s p rou l x, University of British Columbia

j a n e rob e r t s on, University of North Carolina at Chapel Hill

de a n n a c. ru l e , University of Nevada, Las Vegas

m a r l e n e sc hom m e r-a i k i n s, Wichita State University

gr e g ory sc h r aw, University of Nevada, Las Vegas

j u di t h t or n e y-p u r ta, University of Maryland at College Park

c h i n- c h u ng t s a i, National Taiwan University of Science and Technology

l i e v e n v e r sc h a f f e l , University of Leuven

l e a h k. w i l de ng e r, Syracuse University

fa ng -y i ng ya ng, National Taiwan Normal University

j i l l a . z e ru t h, The Pennsylvania State University

Part I

Introduction

3

1 Personal epistemology in the classroom: a welcome and guide for the reader

Florian C. Feucht University of Toledo

Lisa D. Bendixen University of Nevada, Las Vegas

Introduction

Knowledge? One doesn’t need to learn it – one simply knows it. One knows that a tire is round because one can see it. Knowledge sometimes can be heard too. (Amy, age 10)

Actually, one cannot know anything for sure. This is because it has been invented by someone. New inventions can make old ones illogical or false. But one can dis-cuss with other people what they think about it. The numbers and objects have been invented. What does 1 and 1 equal? So one needed to think out what this will be. Then one discussed it at length and decided that it should be named “2.” (Hannah, age 9)

Knowledge is what you know and what you can look up. (Josh, age 12)

I know that knowledge about the woodlands is true by doing experiments that are in my science book and seeing if they come true. Another way to know what is true is you could go with your family during vacations to the woodlands. I am looking at the animals, the plants, their habitat, and watching how they react. By watching what they do you can know if your science book is telling the truth or not. Then I come back and tell it to the teacher. I do this so she can tell it to the class or other classes. I ask her so I can talk about my experiences to the class. (Evan, age 10)

When the Earth originated there were first the bacteria, and the fish, and the dino-saurs. The fish must have known how to swim and the dinosaurs how to walk. This is where knowledge is coming from. (Gwen, age 10)

Knowledge is coming from Greece. That is where they started to write and count… A scholar or a wizard invented the ABCs and the related rules… He told all of this to the king and the king told it to the teacher. The king made the rules for the city, if you don’t stick to the rules, you go to prison. The king wants all people to know what he thinks is interesting. He made a rule that everybody should read and write.

A welcome and guide for the reader4

He also nominated the teacher. One doesn’t need to tell the king everything though. (Linda, age 11)

The previous quotations were made by fourth­grade students from Germany and the US (Feucht, 2008; Haerle, 2006). As can be seen, their views of knowledge and knowing (i.e., personal epistemology) are fascinating, varied, and often linked to educational issues. Personal epistemology and its relevance in the classroom is the topic we pursue in this edited book.

This book incorporates both theoretical and empirical work pertain­ing to personal epistemology (i.e., beliefs about knowledge and know­ing) in the classroom. A large body of theory and research in the field of personal epistemology has been dedicated to college students. Rarely have we addressed the epistemic beliefs of children and young ado­lescents. How it matters in the everyday classroom has also not been investigated thoroughly. Therefore, this book aims to bring together leading­edge research on preschool through secondary students’ per­sonal epistemology and that of their teachers, re­examine existing epistemological frameworks, introduce new models, and provide an empirical foundation for learning and instruction.

Different conceptual frameworks have emerged that define personal epistemology, such as: (1) a developmental progression through different patterns of epistemological thinking (e.g., Baxter Magolda, 1992; King and Kitchener, 1994; Kuhn et al., 2000); (2) epistemological beliefs (Schommer, 1990); (3) epistemological theories (Hofer and Pintrich, 1997); and (4) epistemological resources (Hammer and Elby, 2002).

In an effort to integrate these conceptual frameworks Hofer and Pintrich (1997) define personal epistemology as four identifiable and interrelated dimensions. The first two dimensions describe the nature of knowledge: (1) the certainty of knowledge is focused on the perceived stability and the strength of supporting evidence, and (2) the simplicity of knowledge describes the relative connectedness of knowledge. The remaining two dimensions relate to the process of knowing: (3) the justi­fication of knowledge explains how individuals proceed to evaluate and warrant knowledge claims, and (4) the source of knowledge describes where knowledge resides, internally and/or externally.

A growing body of research provides evidence that personal epistemol­ogy plays a crucial role in the learning of individuals, such as its impact on argumentation, problem­solving, and achievement. The key question of how teachers’ and students’ views of knowledge interact is another area of research that is just beginning to be explored. In addition, per­sonal epistemology’s relation to other fields such as conceptual change,

Florian C. Feucht and Lisa D. Bendixen 5

self­regulated learning, theory of mind, nature of science, motivation, as well as mathematics and science education, is currently emerging.

This book brings together international scholars in the field of per­sonal epistemology. In addition, this group will represent interdisci­plinary perspectives that are key to a more complete and applicable understanding of personal epistemology research.

As the link between personal epistemology and learning and instruc­tion becomes established, the more we will understand the role of per­sonal epistemology in the classroom. Thus, the impetus of the book is to fill a significant gap in our understanding of the relevance of per­sonal epistemology in preschool through secondary education. Finally, we consider the broader implications of this work as it pertains to the importance of personal epistemology and its role in critical thinking development and the education of our future citizens.

The journey to personal epistemology in the classroom

We would like to invite the reader to join our academic journey to per­sonal epistemology in the classroom. In due course, we provide here a brief overview of the terrain/field covered by personal epistemology research since its existence over four decades ago (e.g., Perry, 1970). This brief review does not provide an exhaustive representation of our field. Rather, it aims to provide a contextualized starting point for the reader of this book (see Hofer and Pintrich, 1997, for a more compre­hensive review). We overview a selection of frameworks that have made an important contribution to the field of personal epistemology and that play a crucial role in learning and instruction in the classroom context. We also revisit and consider the frameworks in terms of their meaning for education in general, and personal epistemology in the classroom in particular. In addition, we hope that for the reader new to our field, this will be an effective and efficient base camp for over­looking the existing landscape of personal epistemology and an aid in anticipating what is yet to come in our journey to personal epistemo­logy in the classroom.

Personal epistemology as a developmental trajectory

A variety of frameworks exist that define personal epistemology as a cognitive construct that progresses in its qualities along a predictable developmental path, driven by a process of cognitive disequilibrium (e.g., Baxter Magolda, 1992; Belenky et al., 1986; King and Kitchener, 1994; Kuhn et al., 2000; Perry, 1970). Most notable is Perry’s (1970)

A welcome and guide for the reader6

pioneering Scheme of the intellectual and ethical development of college­aged students. We also briefly summarize the influential models of King and Kitchener (1994) and Kuhn and colleagues (e.g., Kuhn and Weinstock, 2002).

Summarizing the Scheme of the intellectual and ethical development. Perry’s (1970) scheme encompasses nine stages of personal epistemology as intel-lectual and ethical development, which can be summarized in four develop­mental levels. (1) Dualism describes the belief that knowledge and truth are absolute. Individuals at this level hold a polarized, black­ and­white view of the world and label knowledge in clear­cut, right­ and­wrong catego­ries. Authorities are perceived as an omniscient source of knowledge and empowered to administer and communicate knowledge to the learner. In contrast, (2) Multiplicity encompasses the belief that individuals can hold differing knowledge claims. Competing knowledge claims are acceptable (i.e., everyone has the right to be right). Subsequently, the nature of know­ledge is perceived as uncertain and its absolute truth value is doubted. (3) Relativism describes individuals who believe that valid knowledge claims can only be made in relation to their context, such as in a certain domain or era (e.g., history versus science). By using a specific context as a frame of reference some competing knowledge claims are believed to be better than others. (4) Commitment in relativism, the last developmental level, describes the identification of individuals who are certain about the contextualized truth of a knowledge claim but that this is subject to an ongoing process of doubt and refinement.

Summarizing the Reflective judgment model. King and Kitchener (1994, 2002, 2004) developed a framework similar to Perry’s (1970) and fur­ther differentiates its upper levels. The Reflective judgment model defines personal epistemology as epistemic cognition (i.e., a cognitive process superior to, and influential on, both cognition and meta­cognition), which develops along seven stages. These stages can be summarized in three developmental levels: (1) Pre-reflective thinking describes the epis­temic assumption that knowledge is gained through an authoritative figure or through first­hand observation. Knowledge is perceived as absolute and can be known with complete certainty. (2) Quasi-reflective thinking is characterized by the recognition that the certainty of knowl­edge depends on the method of obtaining knowledge. Because aspects of what is known may be wrong or missing, the nature of knowledge is perceived to be ambiguous to a certain extent. (3) Reflective thinking, the final developmental level, encompasses the belief that knowledge is actively constructed and must be evaluated within its context to esti­mate its validity. Knowledge is perceived as uncertain and changing in its nature.

Florian C. Feucht and Lisa D. Bendixen 7

Clearly, the overlap between the developmental levels of Perry’s (1970) scheme and King and Kitchener’s (1994) model is evident; both could be categorized as late­onset developmental models (Chandler et al., 2002). That is, in late adolescence and adulthood and coinciding with the exposure to higher education, the epistemologies of individu­als could progress towards more advanced developmental levels. Until then, individuals, such as elementary and most secondary students, would hold dualistic beliefs about knowledge and/or conduct simple pre­reflective judgments.

Summarizing the Framework of epistemological thinking. Kuhn and col­leagues also propose a developmental model of personal epistemology that considers children as well as adolescents and adults (Kuhn, 1991; Kuhn, et al., 2000; Kuhn and Weinstock, 2002). Their framework defines personal epistemology as epistemological thinking and encom­passes three general developmental levels. At the level of (1) Absolutism a person perceives knowledge as an objective entity, which is located in the external world and can be known with certainty. (2) Multiplism, the contrasting second level, focuses on internalized knowledge source. Subjectivity of knowing and uncertainty of knowledge are important characteristics of this level. At the most advanced level of (3) Evaluativ-ism, both objective and subjective aspects of knowing are incorporated when making knowledge claims. Knowledge is perceived as uncertain, but can be validated within its context.

Revisiting the developmental frameworks. How can epistemic develop­ment frameworks inform personal epistemology in the classroom? In order to assess and foster personal epistemology of learners and teachers alike, these frameworks provide an important theoretical basis for educa­tion as well as teacher training and development. A variety of theoretical and empirical work has proposed that teachers’ personal epistemology, in particular their epistemic development, influences not only their choices of teaching strategies and use of educational materials, but also open­ness to educational reform and further professional development (e.g., Feucht, 2008; Feucht and Bendixen, in press; Patrick and Pintrich, 2001; Schraw and Olafson, 2002; Tsai, 2002). For example, absolutist teachers may tend to perceive teaching as transferring knowledge from teachers as experts to students as naïve and passive learners, while evaluativist teachers may promote learning activities in which students collabora­tively construct knowledge and are expected to justify their knowledge commitments.

As we have stated, both King and Kitchener’s (1994) and Perry’s (1970) work can be considered as late­onset developmental models. That is, the bulk of epistemic development happens during late adolescence

A welcome and guide for the reader8

and adulthood and in combination with learners’ exposure to higher education. Other research, in contrast, has begun to demonstrate that children around the age of ten display some aspects of more advanced epistemic beliefs, such as multiplism and evaluativism (e.g., Feucht, 2008; Haerle, 2006; Mason, 2003). This is consistent with the early­onset understanding of epistemic development (e.g., Chandler et al., 2002). In the context of personal epistemology in the classroom, we see the need for the extension and differentiation of developmental frame­works in terms of providing more detail regarding epistemic develop­ment in earlier ages. For example, the strength of King and Kitchener’s (1994) and Perry’s (1970) models is that they provide great detail in the developmental stages of older students. This developmental preci­sion should also be extended to learners in preschool through second­ary schools. In other words, current developmental models could be revamped and/or new models could be conceptualized that would allow for more systematic and informative recommendations for education.

Education also plays an important role in the epistemic development frameworks. Perry (1970) stresses that it was within the context of higher education, the expectation of independent and critical thinking, and exposure to multiple viewpoints that boosted the epistemic develop­ment of the students he studied. Additionally, he provides suggestions for learning environments that are conducive to epistemic development. Similarly, King and Kitchener (1994) and Kuhn and colleagues (e.g., Kuhn and Weinstock, 2002) consider critical thinking as an important factor in the development of reflective judgment and epistemological thinking, respectively. In general, the experience of critical thinking has become an important consideration of general education in elemen­tary and secondary classrooms as well (e.g., Barnes, 1970; Paul et al., 1990). Therefore, the fostering of critical thinking provides an impor­tant potential for nurturing epistemic development in elementary and secondary learning environments and should not be limited to tertiary education. Following this line of thinking, we encourage the field of personal epistemology to further research and theorize the reciprocal link between epistemic development and critical thinking and the key influence of educational experiences.

While some educational programs and environments can advance epistemic development, others may have a counter productive influence. It has been proposed that certain instructional approaches (e.g., mono­cultures of absolutistic instruction), assessment procedures (e.g., focus on factual knowledge), and/or education in general have the potential to suppress epistemic development (e.g., Chandler, et al., 2002; Feucht, 2008; Walton, 2000). Consistent with our previous comments, we

Florian C. Feucht and Lisa D. Bendixen 9

think that this aspect of educational influence should also be a part of research in personal epistemology development.

Personal epistemology as an epistemological belief system

Summarizing the epistemological belief system. Schommer­Aikins (e.g., Schommer, 1990; Schommer­Aikins, 2002, 2004) conceptualized personal epistemology as a system of more­or­less independent beliefs about knowledge and learning, drawing from a variety of differ­ent research programs (e.g., Perry, 1970; Dweck and Leggett, 1988; Schoenfeld, 1985, 1989). Her framework and its accompanying paper­and­pencil measure enjoys considerable popularity in the field of personal epistemology, and this is evident in the number of studies that are designed around her framework and the application of her measure and the development of its derivations (e.g., Jehng et al., 1993; Schraw et al., 1995; Wood and Kardash, 2002). The framework encompasses five belief dimensions that are proposed to progress in a more asynchronous pattern and are described along the following continua: (1) the structure of knowledge, ranging from discrete to com­plex knowledge; (2) the stability of knowledge, ranging from unchang­ing to evolving; (3) the source of knowledge, ranging from a reliance on authority to observation and reasoning; (4) the speed of learning, ranging from quick or not­at­all learning to gradual learning; and (5) the ability to learn, ranging from innate ability to improvable learn­ing. Due to this conceptualization, Schommer­Aikins’ framework is in contrast to the developmental frameworks previously described that propose more cohesive and predictable levels of personal epis­temology (e.g., King and Kitchener, 1994; Kuhn et al., 2002; Perry, 1970).

Revisiting the epistemological belief system. How can the epistemological belief system inform personal epistemology in the classroom? There are several empirical studies that exist that provide a link between epis­temic beliefs and aspects of academic achievement (e.g., Schommer et al., 1992; Qian and Pan, 2002). For instance, epistemological beliefs have been shown to be associated with science learning (e.g., Bell and Linn, 2002; Elder, 2002), mathematical learning (e.g., DeCorte et al., 2002; Schoenfeld, 1985), cognitive processes (e.g., Kardash and Scholes, 1996; Schommer, 1990), motivation (e.g., Bråten and Strømsø, 2004), and study strategy use (e.g., Schommer et al., 1992; Schrieber and Shinn, 2003).

A small amount of work has also focused on the epistemic beliefs of secondary students and how they are related to achievement in different

A welcome and guide for the reader10

school subjects (e.g., Cano, 2005; Cano and Cardelle­Elawar, 2004; Schommer­Aikins et al., 2005). Such results clearly substantiate the need for more research along these lines looking at the impact of epis­temological beliefs on elementary and secondary school students. For example, are some beliefs more conducive to academic achievement than others? Are some beliefs more important in physics learning than, for example, learning in history? How can teachers better assess stu­dents’ epistemological beliefs to help them instruct and improve the academic achievement of their students?

It is the multidimensional nature of Schommer­Aikins’ framework of epistemological beliefs that is quite intriguing to us and where we see great potential in informing the advancement of epistemic beliefs and to make recommendations for learning and instruction. Differentiat­ing epistemic beliefs along assorted dimensions provides more detailed information on the personal epistemology of preschool, elementary, and secondary school students and their teachers. For example, research­ers could map the epistemic beliefs of students on all five dimensions and/or conduct more in­depth studies on selected dimensions. Such approaches can provide valuable insights into how epistemic beliefs might differ from school subject to school subject. For example, a reli­ance on authority may be an important aspect of knowledge/learning in mathematics while the sources of knowledge in social studies may be more diverse and relative. In addition, are some dimensions not addressed at all in a typical classroom environment? How often are stu­dents required to justify their claims/arguments/opinions in preschool through grade twelve classrooms?

Personal epistemology as epistemological theories

Summarizing the framework of epistemological theories. Hofer and Pintrich (1997, 2002) synthesized a framework on the basis of key aspects iden­tifiable across the fields of personal epistemology and philosophy and define personal epistemology as epistemological theories. This defini­tion incorporates a neo­Piagetian understanding of cognitive develop­ment (Bidell and Fischer, 1992), which is characterized by a more fluid rather than stage­like development, and conceptualizes personal episte­mology as more theory­like rather than as a set of independent beliefs. More specifically, Hofer and Pintrich describe epistemological theories as four identifiable dimensions, that are interrelated and develop in pre­dictable directions: (1) the certainty of knowledge (i.e., stability of knowl­edge and the strength of the supporting evidence); (2) the simplicity of knowledge (i.e., relative connectedness of knowledge); (3) the justification

Florian C. Feucht and Lisa D. Bendixen 11

of knowledge (i.e., procedure to evaluate and warrant knowledge claims); and (4) the source of knowledge (i.e., knowledge resides internally and/or externally). The first and second dimensions describe the nature of knowledge while the third and fourth entail the process of knowing.

Revisiting the framework of epistemological theories. One of the main strengths of Hofer and Pintrich’s framework is that, in our view, it combines a more developmental perspective with a dimensional view of personal epistemology. Considering personal epistemology as theo­ry­like allows the opportunity to research how epistemological beliefs may change while at the same time it permits important detail to be investigated by way of the various dimensions that comprise it. This more complete picture of personal epistemology has great promise and offers guidance for future research as well as many implications for learning and instruction. For example, do different dimensions of epistemological beliefs develop at different rates? Are some dimen­sions more important to learning than others at different ages?

In addition, Hofer and Pintrich’s (1997) conceptualization of per­sonal epistemologies as theory­like overlaps with frameworks of concep­tual change learning (Mason, 2003; Vosniadou, 2003; Vosniadou and Brewer, 1992). This conceptual overlap offers significant opportunity to borrow from the theoretical insights made in the field of conceptual change. For example, the conceptual change literature focuses on the stability of students’ conceptions (i.e., are students’ personal epistemol­ogies in constant flux?), the role of emotions in change, the benefits of intentionality (i.e., the more students are aware of their epistemological theories the more likely substantive change will occur), and processes involved in mechanisms of change.

There are also important educational implications we can draw upon from the conceptual change literature. For instance, in conceptual change learning it is critical for the teacher to assess the prior knowl­edge of their students. In terms of personal epistemology development, assessing epistemic beliefs of students before instruction and lesson planning could also be extremely valuable. In line with the conceptual change principle of offering student’s more elaborate and/or alterna­tive conceptions, teachers could make the epistemology of the content knowledge explicit in their teaching to then challenge students’ exist­ing epistemic beliefs. Do teaching strategies and interventions, such as refutational text and anomalous data found in the conceptual change literature, also apply for personal epistemology change in classroom settings?

A welcome and guide for the reader12

Personal epistemology as epistemological resources

Summarizing the framework of epistemological resources. Hammer and Elby (2002, 2003) propose an epistemological framework that defines per­sonal epistemology within the context of learning in specific subjects such as physics. In essence, students’ personal epistemology is com­prised of a set of fine­grained cognitive resources that are catego rized in four areas: (1) nature and sources of knowledge (e.g., knowledge as propa­gated stuff, knowledge as free creation, and knowledge as fabricated stuff); (2) epistemological activities (e.g., accumulation, formation, and checking); (3) epistemological forms (e.g., stories, rules, facts, and games); and (4) epistemological stances (e.g., acceptance, understand­ing, and puzzlement). These epistemological resources are activated by, sensitive to, and dependent upon the context of each individual and are not necessarily subject to a developmental progression.

Revisiting the framework of epistemological resources. A clear strength of Hammer and Elby’s (2002, 2003) conceptualization of epistemological resources is that it stems from classroom research on children and young adolescents, and therein, applies directly to understanding personal epistemology in the classroom. For example, their situated approach to researching personal epistemology within specific school subjects and the fact that it is anchored even within particular lesson plans allows them to derive specific, concrete, and teacher­friendly epistemic recom­mendations that are conducive to students’ learning.

Although Elby and Hammer’s goal is not to generalize too far beyond individual classrooms, we see valuable educational implications in their work. We see this approach as fundamental in establishing portfolios of varied instructional practices (i.e., a bag of epistemic tricks) that can be used to influence students’ personal epistemology and, therein, to strategically foster their learning. The transferability of such pedagogi­cal practices across different contexts and school subjects could also become part of more experimental research designs. For example, the methodological question remains as to what kind of categories can be used to differentiate between epistemic resources that are more condu­cive to the learning of content knowledge than others and to distinguish between educational contexts that are more influential on epistemo­logical resources.

Organization of the book

We decided to organize the book into five parts: (I) Introduction; (II) Frameworks and conceptual issues; (III) Students’ personal epistemology,

Florian C. Feucht and Lisa D. Bendixen 13

its development, and its relation to learning; (IV) Teachers’ personal epis-temology and its impact on classroom teaching; and (V) Conclusion. We provide a brief rationale for the scope of each part of the book and introduce its contributing authors along with a succinct description of their chapter’s content. In addition, we developed an advanced organizer to provide the reader with a basic overview of the themes and highlights featured in the book (see Table 1.1). Overall, our aim is to provide a guiding structure to the book based on the main focus of each chapter. Many chapters actually address aspects that are of relevance to multiple parts of the book (i.e., educational issues, stu­dents, and teachers), which illustrates the interrelatedness of the top­ics featured in this book. To aid in navigating this complexity, we also highlight general themes and topics that cut across the organization of the book in the form of Important things to look for.

Framing the book: introduction and conclusion

The first and last sections of the book are written by the editors to frame the main chapters for the reader. The introduction (Chapter 1) provides our rationale and goals for the book, a brief review of exist­ing frameworks in the field of personal epistemology, and a discussion of important educational issues and questions to bear in mind when reading the book. The idea is to provide a foundation that will help the reader in contextualizing the conceptual, empirical, and method­ological similarities, differences, and innovations put forward by the contributing chapter authors along with their discussion of related edu­cational implications.

The goal of the conclusion (Chapter 17) is to highlight, discuss, and bring together emerging issues of personal epistemology addressed in the preceding chapters with a focus on innovations regarding personal epistemology in the classroom. Essentially, the conclusion aims to pro­vide a synthesized discussion of what we know about personal episte­mology in the classroom as a scientific community, to outline ideas to advance theory, research, and educational practice beyond the book, and, therein, to push the field of personal epistemology further.

Part II: Frameworks and conceptual issues

The second section of the book includes chapters that broadly consider how personal epistemology matters in the classroom. The chapter con­tributors of this section address general aspects such as the development

Tab

le 1

.1. O

verv

iew

of t

he b

ook

Ch

apte

r

Tit

le a

nd

auth

or(s

)

Foc

us

of

chap

ter

Per

son

al

epis

tem

olog

y d

escr

ibed

as

…

Dom

ain

­gen

eral

an

d/o

r do

mai

n­s

peci

fic

Pop

ula

tion

Ad

dit

ion

al

issu

es a

dd

ress

ed

Part

I: In

trod

uctio

n

1P

erso

nal e

pist

emol

ogy

in

the

clas

sroo

m: a

wel

com

e an

d gu

ide

for

the

read

er

Feuc

ht a

nd B

endi

xen

Con

cept

ual

Per

sona

l ep

iste

mol

ogy

Dom

ain­

gene

ral

and

dom

ain­

spec

ific

Pre

scho

ol t

o gr

ade

twel

ve

stud

ents

and

te

ache

rs

Exi

stin

g fr

amew

orks

; ov

ervi

ew a

nd g

uide

fo

r th

e bo

ok

Part

II: F

ram

ewor

ks a

nd c

once

ptua

l iss

ues

2M

anif

esta

tion

s of

an

epis

tem

olog

ical

bel

ief

syst

em in

pre

scho

ol t

o gr

ade

twel

ve c

lass

room

s S

chom

mer

-Aik

ins,

Bir

d,

and

Bak

ken

Con

cept

ual

Epi

stem

olog

ical

be

lief

syst

emD

omai

n­ge

nera

l an

d sp

ecifi

c sc

ienc

e an

d hi

stor

y

Pre

scho

ol t

o gr

ade

twel

ve

stud

ents

, te

ache

rs, a

nd

pare

nts

Lon

gitu

dina

l ex

ampl

es;

deve

lopm

enta

l re

curs

ion;

soc

io­

cult

ural

per

spec

tive

3E

pist

emic

clim

ate

in

elem

enta

ry c

lass

room

s Fe

ucht

Con

cept

ual

fram

ewor

kP

erso

nal

epis

tem

olog

yD

omai

n­ge

nera

l an

d sp

ecifi

c;

scie

nce

and

hist

ory

Ele

men

tary

sc

hool

st

uden

ts a

nd

teac

hers

Epi

stem

ic c

limat

e;

epis

tem

olog

y of

inst

ruct

ion

and

educ

atio

nal

mat

eria

ls

4T

he in

tegr

ativ

e m

odel

of

pers

onal

epi

stem

olog

y de

velo

pmen

t: t

heor

etic

al

unde

rpin

ning

s an

d im

plic

atio

ns f

or e

duca

tion

R

ule

and

Ben

dixe

n

Con

cept

ual

Per

sona

l ep

iste

mol

ogy

Dom

ain­

gene

ral

and

spec

ific;

m

athe

mat

ics

Ele

men

tary

sc

hool

st

uden

ts a

nd

teac

hers

The

oret

ical

ex

amin

atio

n of

m

odel

; epi

stem

ic

volit

ion;

edu

cati

onal

im

plic

atio

ns

Ch

apte

r

Tit

le a

nd

auth

or(s

)

Foc

us

of

chap

ter

Per

son

al

epis

tem

olog

y d

escr

ibed

as

…

Dom

ain

­gen

eral

an

d/o

r do

mai

n­s

peci

fic

Pop

ula

tion

Ad

dit

ion

al

issu

es a

dd

ress

ed

5A

n ep

iste

mic

fra

mew

ork

for

scie

ntifi

c re

ason

ing

in

info

rmal

con

text

sYa

ng a

nd T

sai

Em

piri

cal

fram

ewor

kP

erso

nal

epis

tem

olog

yD

omai

n­sp

ecifi

c;

info

rmal

sci

ence

Ele

men

tary

, m

iddl

e, a

nd

high

sch

ool

stud

ents

Sci

enti

fic r

easo

ning

in

info

rmal

co

ntex

ts; e

pist

emic

de

velo

pmen

t

6W

ho k

now

s w

hat

and

who

can

we

belie

ve?

Epi

stem

olog

ical

bel

iefs

ar

e be

liefs

abo

ut

know

ledg

e (m

ostl

y) t

o be

att

aine

d fr

om o

ther

s B

rom

me,

Kie

nhue

s, an

d Po

rsch

Con

cept

ual

and

empi

rica

l

Epi

stem

olog

ical

be

liefs

Dom

ain­

gene

ral

and

spec

ific,

m

athe

mat

ics

and

scie

nce

Ele

men

tary

sc

hool

st

uden

ts

Div

isio

n of

cog

niti

ve

labo

r; s

econ

d­ha

nd k

now

ledg

e ev

alua

tion

Par

t II

I: S

tud

ents

’ per

son

al e

pis

tem

olog

y, i

ts d

evel

opm

ent,

an

d i

ts r

elat

ion

to

lear

nin

g

7S

talk

ing

youn

g pe

rson

s’

chan

ging

bel

iefs

abo

ut

belie

fC

hand

ler

and

Pro

ulx

Con

cept

ual

Fol

k ep

iste

mol

ogy;

pe

rson

al

epis

tem

olog

y

Dom

ain­

gene

ral

Ear

ly

child

hood

Dis

tinc

tion

bet

wee

n pe

rson

al, t

aste

/bru

te

fact

s, a

nd s

ocia

l fa

cts/

shar

ed v

alue

s

8

Epi

stem

olog

ical

de

velo

pmen

t in

ver

y yo

ung

know

ers

W

ilden

ger,

Hof

er, a

nd B

urr

Con

cept

ual

and

empi

rica

l

Per

sona

l ep

iste

mol

ogy;

ep

iste

mol

ogic

al

thou

ght

Dom

ain­

spec

ific;

ta

ste,

am

bigu

ous

fact

, mor

alit

y, a

nd

fact

Pre

scho

ol

stud

ents

Ori

gins

of

epis

tem

olog

ical

de

velo

pmen

t;

egoc

entr

ic

subj

ecti

vity

; the

ory

of m

ind

Ch

apte

r

Tit

le a

nd

auth

or(s

)

Foc

us

of

chap

ter

Per

son

al

epis

tem

olog

y d

escr

ibed

as

…

Dom

ain

­gen

eral

an

d/o

r do

mai

n­s

peci

fic

Pop

ula

tion

Ad

dit

ion

al

issu

es a

dd

ress

ed

9B

elie

fs a

bout

kno

wle

dge

and

revi

sion

of k

now

ledg

e:

on t

he im

port

ance

of

epis

tem

ic b

elie

fs fo

r in

tent

iona

l con

cept

ual

chan

ge in

ele

men

tary

and

m

iddl

e sc

hool

stu

dent

sM

ason

Con

cept

ual

and

empr

icia

l

Epi

stem

ic

belie

fsD

omai

n­sp

ecifi

c;

scie

nce

and

hist

ory

Ele

men

tary

an

d m

iddl

e sc

hool

st

uden

ts

Epi

stem

ic b

elie

fs

impa

ctin

g co

ncep

tual

cha

nge;

af

fect

; sch

ool

tran

siti

ons;

sci

ence

ed

ucat

ion

10T

he r

eflex

ive

rela

tion

be

twee

n st

uden

ts’

mat

hem

atic

s­re

late

d be

liefs

and

the

m

athe

mat

ics

clas

sroo

m

cult

ure

De

Cor

te, O

p ’t

Eyn

de,

Dep

aepe

, and

Ver

scha

ffel

Con

cept

ual

and

empi

rica

l

Mat

hem

atic

s­re

late

d be

liefs

; ep

iste

mol

ogic

al

belie

fs a

bout

m

athe

mat

ics;

be

liefs

abo

ut

mat

hem

atic

s le

arni

ng

Dom

ain­

spec

ific,

m

athe

mat

ics

Ele

men

tary

an

d se

cond

ary

scho

ol

stud

ents

and

te

ache

rs

Bel

iefs

abo

ut

mat

hem

atic

s di

spos

itio

n;

clas

sroo

m c

ultu

re;

soci

o­co

nstr

ucti

vist

pe

rspe

ctiv

e

11E

xam

inin

g th

e in

fluen

ce

of e

pist

emic

bel

iefs

and

go

al o

rien

tati

ons

on t

he

acad

emic

per

form

ance

of

ado

lesc

ent

stud

ents

en

rolle

d in

hig

h­po

vert

y,

high

­min

orit

y sc

hool

s M

urph

y, B

uehl

, Zer

uth,

E

dwar

ds, L

ong,

and

Mon

oi

Em

piri

cal

Epi

stem

ic

belie

fsD

omai

n­ge

nera

l; be

liefs

abo

ut

lear

ning

and

w

ork

Mid

dle

and

high

sch

ool

stud

ents

Hig

h­m

inor

ity

and

high

­pov

erty

sc

hool

s; e

pist

emic

de

velo

pmen

t; s

choo

l tr

ansi

tion

s

Tab

le 1

.1. (

cont

.)

Ch

apte

r

Tit

le a

nd

auth

or(s

)

Foc

us

of

chap

ter

Per

son

al

epis

tem

olog

y d

escr

ibed

as

…

Dom

ain

­gen

eral

an

d/o

r do

mai

n­s

peci

fic

Pop

ula

tion

Ad

dit

ion

al

issu

es a

dd

ress

ed

12U

sing

cog

niti

ve

inte

rvie

win

g to

ex

plor

e el

emen

tary

an

d se

cond

ary

scho

ol

stud

ents

’ epi

stem

ic a

nd

onto

logi

cal c

ogni

tion

G

reen

e, T

orne

y-P

urta

, A

zeve

do, a

nd R

ober

tson

Con

cept

ual,

met

hodo

logi

cal

and

empi

rica

l

Epi

stem

ic a

nd

onto

logi

cal

cogn

itio

n

Dom

ain­

spec

ific

mat

hem

atic

s;

phys

ics,

his

tory

, an

d po

litic

al

scie

nce

Ele

men

tary

an

d se

cond

ary

scho

ol

stud

ents

Eps

item

ic a

nd

onto

logi

cal

cogn

itio

n;

mea

sure

men

t va

lidat

ion;

epi

stem

ic

deve

lopm

ent

Par

t IV

: Tea

cher

s’ p

erso

nal

ep

iste

mol

ogy

and

its

im

pac

t on

cla

ssro

om t

each

ing

13E

pist

emol

ogic

al r

esou

rces

an

d fr

amin

g: a

cog

niti

ve

fram

ewor

k fo

r he

lpin

g te

ache

rs in

terp

ret

and

resp

ond

to t

heir

stu

dent

s’

epis

tem

olog

ies

Elb

y an

d H

amm

er

Con

cept

ual

fram

ewor

kE

pist

emol

ogic

al

reso

urce

s an

d fr

amin

g

Dom

ain­

spec

ific,

phy

sics

an

d te

ache

rs’

prof

essi

onal

kn

owle

dge

Ele

men

tary

sc

hool

tea

cher

s an

d st

uden

ts

Cla

ssro

om c

onte

xt;

epis

tem

ic f

ram

ing

of s

tude

nts;

and

in

stru

ctio

nal

impl

icat

ions

14T

he e

ffec

ts o

f te

ache

rs’ b

elie

fs o

n el

emen

tary

stu

dent

s’

belie

fs, m

otiv

atio

n,

and

achi

evem

ent

in

mat

hem

atic

sM

uis

and

Foy

Em

piri

cal

Epi

stem

ic b

elie

fsD

omai

n­sp

ecifi

c;

mat

hem

atic

sE

lem

enta

ry

scho

ol t

each

ers

and

stud

ents

Dir

ect

exam

inat

ion

of t

each

ers’

bel

iefs

on

stu

dent

bel

iefs

, se

lf­e

ffica

cy, a

nd

goal

s; p

ath

anal

ysis

15

Tea

cher

s’ a

rtic

ulat

ion

of

belie

fs a

bout

tea

chin

g kn

owle

dge:

co

ncep

tual

izin

g

a be

lief

fram

ewor

k

Fiv

es a

nd B

uehl

Con

cept

ual a

nd

empi

rica

l

Bel

iefs

abo

ut

teac

hing

kn

owle

dge;

ep

iste

mic

bel

iefs

Dom

ain­

spec

ific,

te

achi

ng

know

ledg

e

Pre

serv

ice

and

inse

rvic

e el

emen

tary

, m

iddl

e, a

nd

high

sch

ool

teac

hers

Tea

chin

g kn

owle

dge

as a

dom

ain;

im

plic

atio

ns f

or

teac

her

educ

atio

n

Ch

apte

r

Tit

le a

nd

auth

or(s

)

Foc

us

of

chap

ter

Per

son

al

epis

tem

olog

y d

escr

ibed

as

…

Dom

ain

­gen

eral

an

d/o

r do

mai

n­s

peci

fic

Pop

ula

tion

Ad

dit

ion

al

issu

es a

dd

ress

ed

16B

eyon

d ep

iste

mol

ogy:

ass

essi

ng

teac

hers

’ epi

stem

olog

ical

an

d on

tolo

gica

l wor

ldvi

ews

Ola

fson

and

Sch

raw

Met

hodo

logi

cal

and

empi

rica

lE

pist

emol

ogic

al

and

onto

logi

cal

wor

ldvi

ews

Dom

ain­

gene

ral

Inse

rvic

e te

ache

rs

enro

lled

in

grad

uate

sc

hool

Ont

olog

ical

and

ep

iste

mol

ogic

al

disc

tinc

tion

; te

achi

ng v

igne

ttes

; fo

ur­q

uadr

ant

scal

e;

teac

her

belie

fs a

nd

prac

tice

Par

t V: C

oncl

usi

on

17

Per

sona

l epi

stem

olog

y in

th

e cl

assr

oom

: wha

t do

es

rese

arch

and

the

ory

tell

us

and

whe

re d

o w

e ne

ed t

o go

nex

t?

Ben

dixe

n an

d Fe

ucht

Con

cept

ual

Per

sona

l ep

iste

mol

ogy

Dom

ain­

gene

ral

and

dom

ain­

spec

ific

Pre

scho

ol t

o gr

ade

twel

ve

stud

ents

and

te

ache

rs

The

mes

and

hi

ghlig

hts

disc

usse

d;

impl

icat

ions

for

le

arni

ng a

nd

inst

ruct

ion;

gen

eral

ed

ucat

iona

l goa

ls

Tab

le 1

.1. (

cont

.)

Florian C. Feucht and Lisa D. Bendixen 19

of epistemic beliefs through education, but also focus on the roles of, and interrelationships among, students, peers, teachers, and parents. Schommer­Aikins, Bird, and Bakken (Chapter 2) describe the mani­festation of an epistemological belief system in preschool to grade twelve classrooms and consider epistemic development from systemic and socio­constructivist perspectives. Next, Feucht (Chapter 3) con­ceptualizes an educational model defining epistemic climate, a phenom­enon stemming from epistemological characteristics found in students, teachers, instruction, and educational materials. Rule and Bendixen (Chapter 4) examine the conceptual roots of their model and its educa­tional implications. In the next chapter, Yang and Tsai (Chapter 5) pro­pose a theoretical framework based on several of their empirical studies that conceptualizes the development and interrelation of scientific rea­soning and personal epistemology in informal contexts.

In the last chapter of this section, Bromme, Kienhues, and Porsch (Chapter 6) discuss the division of cognitive labor, its role in the acquisition of knowledge, and how it matters to life­long learning and the development of good citizenship. All contributors follow a more domain­general approach to personal epistemology and discuss useful and practical suggestions for enhancing teaching and learning in the classroom.

Part III: Students’ personal epistemology, its development, and its relation to learning

In the third section, general views of students’ personal epistemology are considered. These chapter authors present leading­edge research, reconsider conceptual issues and frameworks in the field, and address methodological problems that occur when dealing with personal epis­temology in the contexts of preschool, elementary, and/or secondary education. Important issues include epistemic development, classroom contexts, motivation, and specific student populations. Most of the contributing authors in this section focus on domain/content­specific aspects of personal epistemology. The chapters are ordered accord­ing to the age of the students being considered. Chandler and Proulx (Chapter 7) conceptually rethink the development of personal episte­mologies in early childhood and argue for a focus on social facts and shared values (rather than brute facts and personal taste) to more accu­rately capture personal epistemology. Looking at a similar age group, Wildenger, Hofer, and Burr (Chapter 8) explore empirically the origins of epistemological development, its links to theory of mind, and educa­tional implications in the preschool and Kindergarten context. In the

A welcome and guide for the reader20

following chapter, Mason (Chapter 9) embarks upon the link between epistemic beliefs and intentional conceptual change learning in ele­mentary and middle school students and points out the importance of cognitive, emotional, and contextual factors in her discussion.

From a socio­constructivist perspective, De Corte, Op ’t Eynde, Depaepe, and Verschaffel (Chapter 10) provide an empirical and theoretical discussion on the ref lexive relation between students’ mathematics­related beliefs and classroom culture and promote the need to develop positive mathematical dispositions early in educa­tion. Murphy, Buehl, Zeruth, Edwards, Long, and Monoi (Chapter 11) examine the inf luence of epistemic beliefs and goal orientation on the academic performance of adolescent students in high­poverty, high­minority schools. Greene, Torney­Purta, Azevedo, and Rob­ertson (Chapter 12) use cognitive and semi­structured interviewing techniques to methodologically verify items in their newly­developed measure and to tap the epistemic and ontological cognition of elemen­tary and secondary school students. The latter two chapters address aspects of epistemic development and explore differences between various school subjects. All contributors provide suggestions for developmentally appropriate instruction for preschool through sec­ondary education.

Part IV: Teachers’ personal epistemology and its impact on classroom teaching

The fourth section of the book focuses on research that examines teachers’ personal epistemology and its impact on student learning, instructional practice, and teacher development. While all chapter authors present empirical and conceptual issues related to teachers’ personal epistemology and pedagogical knowledge, each chapter has a specific focus. Within the context of physics education, Elby and Hammer (Chapter 13) elaborate on their cognitive framework of epis­temic resources and frames and discuss how it can be used by teachers to interpret and respond to their students’ epistemologies. Muis and Foy (Chapter 14) explore the effects of teachers’ beliefs on elemen­tary students’ beliefs, motivation, and achievement in mathematics. The following two chapters focus on the exploration and implica­tions of teachers’ epistemological beliefs. Fives and Buehl (Chapter 15) explore the articulation of beliefs about teaching knowledge, con­ceptualize a multi­faceted belief framework, and discuss aspects of belief change in teacher education. Olafson and Schraw (Chapter 16)

Florian C. Feucht and Lisa D. Bendixen 21

assess the epistemological and ontological worldviews of teachers in relation to curriculum and instruction, applying a mixed methodo­logical approach using their four­quadrant scale. All authors in this section put forward conceptual ideas and educational recommenda­tions to inform theory, research, practice, and teacher education.

Important things to look for in this book

A closer examination of the chapters reveals important topics and themes that are crucial to the field of personal epistemology, and these may not be explicitly reflected in the sections we use to organize the book. With this in mind, we would like to point out for the reader some additional important things to look for.

Look for: insights into the construct of personal epistemology

The field of personal epistemology research has nourished a large amount of different definitions of what personal epistemology is and what it is not. Despite Hofer and Pintrich’s (1997, 2002) efforts to estab­lish an overarching framework to encompass the existing viewpoints on the construct of personal epistemology, we seem to have a long way to go toward consensus. We are excited to see that significant ground has been gained along these lines in the coming chapters. As the reader will see, many of the authors are beginning to speak in more common terms and are acknowledging the similarities that do exist in what was once considered disparate perspectives. We think this is a tremendous step forward for the field.

Look for: new and extended frameworks of personal epistemology

One of the goals of the book is to provide a platform to rethink and/or elaborate on existing frameworks in the field of personal epistemology and, certainly, to pose new conceptualizations in light of education. Several authors have revamped their existing frameworks and others offer new ones that pay closer attention to educational implications and practice. The advantages of frameworks that attempt to clearly cap­ture the essence of personal epistemology are numerous. We see the frameworks that are described in this volume as not only conceptually valuable but as guides that can be used successfully by researchers and educators alike.

A welcome and guide for the reader22

Look for: domain-general and domain-specific aspects of personal epistemology

Whether or not personal epistemology is domain­general or domain­specific is a question that we have grappled with in our field for a number of years (e.g., Alexander, 2006; Buehl et al., 2002; Hofer, 2006; Muis et al., 2006). In our view, the work represented in this volume has moved beyond this simple question to more nuanced discussions and investigations. For example, you will see that the boundaries of what is meant by “domain” will be expanded (e.g., various school subjects, content areas, and teacher knowledge) and the notion of how domain­general and domain­specific epistemological beliefs may work together will be considered. Again, how these aspects of personal epistemology impact learning and instruction will be offered.

Look for: epistemic development and mechanisms of change

Several developmental frameworks exist that describe the maturation of epistemological beliefs during adolescence and adulthood (Baxter Magolda, 1992; Belenky et al., 1986; Perry, 1970; King and Kitchener, 1994). Little is known about epistemic development in childhood and early adolescence and the mechanisms that bring about change (Bendixen and Rule, 2004; Burr and Hofer, 2002; Haerle, 2006). Sev­eral chapters in this book address these important issues in school­age children and adolescents. Educational practice as a mechanism for epistemic development in and of itself will be elaborated upon with insightful and realistic recommendations for teachers and students as well.

Look for: epistemic climate and classroom cultures

That personal epistemologies might be influenced through processes of enculturation (Vygotsky, 1978) and/or systemic processes (Bron­fenbrenner, 1979) has only recently been considered in any great detail (e.g., Bendixen and Rule, 2004; Haerle and Bendixen, 2008; Schommer­Aikins, 2004). The epistemic underpinnings and messages about knowledge and knowing that abound in classrooms will be viv­idly described and researched by a number of the chapter authors. This clarity in how personal epistemology operates within the classroom culture/community will be presented as being extremely valuable for future research endeavors and educational practice.

Florian C. Feucht and Lisa D. Bendixen 23

Look for: advancement in measurement issues

The valid and reliable measurement of personal epistemology has proven to be an arduous mission in its own right. Over the years, researchers have called for more diverse methodological approaches to address the complexity we know exists within the construct (e.g., Bendixen and Rule, 2004; Hofer and Pintrich, 1997; Pintrich, 2002). Recently, we have asked, how can personal epistemology be explored and measured in school settings that push the research field forward and inform edu­cational practice at the same time (Haerle and Bendixen, 2008)? The contributors in this book have certainly risen to the occasion in terms of addressing this difficult yet practical question.

Personal epistemology: the individual, education, and society

It is remarkable to us how the writings of William Perry (1970) still ring true today and we would like to acknowledge this as we consider our final thoughts regarding our goals for this book. Although his developmental scheme and research did not directly address the edu­cational implications of personal epistemology, Perry certainly saw its tremendous potential. Although the potential is there, Perry also states quite firmly that society demands that individuals be equipped to make commitments within our relative world and that the educational com­munity must be there for the student if they are to progress epistemo­logically. In his view, the system of education at the time was not in line with these demands:

… we feel that educational mores have not kept up with this century’s changes in the nature of knowledge or with the demands the new relativism places upon the learner. (p. 213)

Similarly, in referencing the findings from his landmark study, he states that

we hear the students as hungering for a nutriment essential to growth and meagerly supplied within the conventions of present­day education. The growth demanded of them, and for which they yearn, involves a new kind of responsibility. (p. 214)

As we embark on this new and complex twenty­first century, the need Perry refers to in terms of the demands of a relativistic society seems to be even greater. We see that part of this new responsibility resides in

A welcome and guide for the reader24

theory, research, and practice dedicated to personal epistemology and to the education of the future citizens of our world: the students.

The field of personal epistemology is poised to make significant con­tributions to the individual, to education, and to society. Of course, these three components are intricately tied to one another and we feel strongly that personal epistemology theory and research can bridge them all and advance them in various ways. Because of this, we have asked all of the authors in this volume to take a serious and thorough look at how their work contributes to education and, as the reader will see, they have certainly done so with resounding success.

A final welcome

We hope this introduction has provided some valuable context and insight for the reader as they consider the leading­edge scholarship accu­mulated in this volume. We also hope that the following chapters are as insightful and inspiring for the reader as they are for us. Enjoy the jour­ney and, again, welcome to personal epistemology in the classroom.

R EF ER ENCES

Alexander, P. A. (2006). What would Dewey say? Channeling Dewey on the issue of specificity of epistemic beliefs: A response to Muis, Bendixen, and Haerle (2006). Educational Psychology Review, 18(1), 55–65.

Barnes, D. L. (1970). Identifying and using critical thinking skills in the elementary classroom. (ERIC Document Reproduction Service No. EJ034804).

Baxter Magolda, M. B. (1992). Knowing and reasoning in college: Gender-related patterns in students’ intellectual development. San Francisco: Jossey­Bass.

Belenky, M. F., Clinchy, B. M., Goldberger, N. R., and Tarule, J. M. (1986). Women’s ways of knowing: The development of self, voice, and mind. New York: Basic Books.

Bell, P. and Linn, M. C. (2002). Beliefs about science: How does science instruction contribute? In B. K. Hofer and P. R. Pintrich (Eds.), The psy-chology of beliefs about knowledge and knowing (321–46). Mahwah, NJ: Law­rence Erlbaum Associates.

Bendixen, L. D. (2002). A process model of epistemic belief change. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (191–208). Mahwah, NJ: Lawrence Erlbaum Associates.

Bendixen, L. D. and Rule, D. C. (2004). An integrative approach to personal epistemology: A guiding model. Educational Psychologist, 39, 69–80.

Bidell, T. R. and Fischer, K. (1992). Beyond the stage debate: Action, structure, and variability in Piagetian theory and research. In R. J. Sternberg and C. A. Berg (Eds.), Intellectual development (100–40). New York: Cambridge University Press.

Florian C. Feucht and Lisa D. Bendixen 25

Bråten, I. and Strømsø, H. I. (2004). Epistemological beliefs and implicit theo­ries of intelligence as predictors of achievement goals. Contemporary Edu-cational Psychology, 29(4), 371–88.

Bronfenbrenner, U. (1979). The ecology of human development. Cambridge, MA: Harvard University Press.

Buehl, M. M., Alexander, P. A., and Murphy, P. K. (2002). Beliefs about schooled knowledge: Domain general or domain specific? Contemporary Educational Psychology, 27, 415–49.

Burr, J. E. and Hofer, B. K. (2002). Personal epistemology and theory of mind: Deciphering young children’s beliefs about knowledge and know­ing. New Ideas in Psychology, 20, 199–224.

Cano, F. (2005). Epistemological beliefs and approaches to learning: Their change through secondary school and their influence on academic per­formance. British Journal of Educational Psychology, 75, 203–21.

Cano, F. and Cardelle­Elawar, M. (2004). An integrated analysis of secondary school students’ conceptions and beliefs about learning. European Journal of Psychology of Education, 19(2), 167–87.

Chandler, M. J., Hallett, D., and Sokol, B. W. (2002). Competing claims about competing knowledge claims. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (154–68). Mahwah, NJ: Lawrence Erlbaum Associates.

De Corte, E., Op ’t Eynde, P., and Verschaffel, L. (2002). “Knowing what to believe”: the relevance of students’ mathematical beliefs for mathemat­ics education. In B. K. Hofer and P. R. Pintrich (Eds.), Personal episte-mology: The psychology of beliefs about knowledge and knowing (297–320). Mahwah, NJ: Lawrence Erlbaum Associates.

Dweck, C. S. and Leggett, E. L. (1988). A social­cognitive approach to motiva­tion and personality. Psychological Review, 95(2), 256–73.

Elder, A. D. (2002). Characterizing fifth grade students’ epistemological beliefs in science. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

Feucht, F. C. (2008). The nature of epistemic climates in elementary classrooms. Unpublished doctoral dissertation. University of Nevada, Las Vegas.

Feucht, F. C. and Bendixen, L. D. (in press). Exploring similarities and differ­ences in personal epistemologies of US and German elementary school teachers. Cognition and Instruction.

Haerle, F. C. (2006). Personal epistemologies of fourth-graders: Their beliefs about knowledge and knowing. Oldenburg, Germany: Didaktisches Zentrum.

Haerle, F. C. and Bendixen, L. D. (2008). Personal epistemology in elementary classrooms: A conceptual comparison of Germany and the United States and a guide for future cross­cultural research. In M. S. Khine (Ed.), Knowing, knowledge and beliefs: Epistemological studies across diverse cultures (151–76). New York, NY: Springer.

Hammer, D. H. and Elby, A. (2002). On the form of personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology

A welcome and guide for the reader26

of beliefs about knowledge and knowing (169–90). Mahwah, NJ: Lawrence Erlbaum Associates.

(2003). Tapping epistemological resources for learning physics. Journal of the Learning Sciences, 12(1), 53–90.

Hofer, B. K. (2006). Beliefs about knowledge and knowing: Integrating domain specificity and domain generality. Educational Psychology Review, 18, 67–76.

Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learn­ing. Review of Educational Research, 67, 88–140.

(2002). Personal epistemology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

Jehng, J. J., Johnson, S. D., and Anderson, R. C. (1993). Schooling and stu­dents’ epistemological beliefs about learning. Contemporary Educational Psychology, 18, 23–35.

Kardash, C. M. and Scholes, R. J. (1996). Effects of preexisting beliefs, episte­mological beliefs, and need for cognition on interpretation of controversial issues. Journal of Educational Psychology, 88, 260–71.

King, P. M. and Kitchener, K. S. (1994). Developing reflective judgment: Under-standing and promoting intellectual growth and critical thinking in adolescents and adults. San Francisco: Jossey­Bass.

(2002). The reflective judgment model: Twenty years of research on epis­temic cognition. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing (37–61). Mahwah, NJ: Lawrence Erlbaum Associates.

(2004). Reflective judgment: Theory and research on the development of epistemic assumptions. Educational Psychologist, 39(1), 5–18.

Kuhn, D. (1991). The skills of argument. New York: Cambridge University Press.Kuhn, D., Cheney, R., and Weinstock, M. (2000). The development of episte­

mological understanding. Cognitive Development, 15, 309–28.Kuhn, D. and Weinstock, M. (2002). What is epistemological thinking and

why does it matter? In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing (121–44). Mahwah, NJ: Lawrence Erlbaum Associates.

Lidar, M., Lundqvist, E., and Ostman, L. (2005). The interplay between teachers’ epistemological moves and students’ practical epistemology. Sci-ence Education, 89, 149–63.

Mason, L. (2003). Personal epistemologies and intentional conceptual change. In G. M. Sinatra and P. R. Pintrich (Eds.), Intentional conceptual change (199–236). Mahwah, NJ: Lawrence Erlbaum Associates.

Muis, K. R., Bendixen, L. D., and Haerle, F. C. (2006). Domain­generality and domain­specificity in personal epistemology research: Philosophical and empirical reflections in the development of a theoretical framework. Educational Psychology Review, 18(1), 3–54.

Patrick, H. and Pintrich, P. R. (2001). Conceptual change in teachers’ intui­tive conceptions of learning, motivation, and instructions: The role of motivational and epistemological beliefs. In B. Torf and R. Sternberg (Eds.), Understanding and teaching the intuitive mind (117–43). Mahwah, NJ: Lawrence Erlbaum Associates.

Florian C. Feucht and Lisa D. Bendixen 27

Paul, R. W., Binker, A. J., Jensen, K., and Kreklau, H. (1990). Critical think-ing handbook: 4th–6th grades. Dillon Beach, CA: Foundation for Critical Thinking.

Perry, W. G. Jr. (1970). Forms of intellectual and ethical development in the college years: A scheme. New York: Holt, Rinehart, and Winston.

Pintrich, P. R. (2002). Future challenges and directions for theory and research on personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Per-sonal epistemology: The psychology of beliefs about knowledge and knowing (389–414). Mahwah, NJ: Lawrence Erlbaum Associates.

Qian, G. and Pan, J. (2002). A comparison of epistemological beliefs and learn­ing from science text between American and Chinese high school students. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychol-ogy of beliefs about knowledge and knowing (365–85). Mahwah, NJ: Law­rence Erlbaum Associates.

Schoenfeld, A. H. (1985). Mathematical problem solving. New York: Academic. (1989). Exploration of students’ mathematical beliefs and behavior. Journal

for Research in Mathematics Education, 20, 338–55.Schommer, M. (1990). Effects of beliefs about the nature of knowledge on

comprehension. Journal of Educational Psychology, 82, 498–504.Schommer, M., Crouse, A., and Rhodes, N. (1992). Epistemological beliefs

and mathematical text comprehension: Believing it is simple does not make it so. Journal of Educational Psychology, 84(4), 435–43.

Schommer­Aikins, M. (2002). An evolving theoretical framework for an epis­temological belief system. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (103–18). Mahwah, NJ: Lawrence Erlbaum Associates.

(2004). A systemic approach to the conceptualization and study of epistemo­logical beliefs: When researchers coordinate and cooperate. Educational Psychologist, 39, 19–30.

Schommer­Aikins, M., Duell, O., and Hutter, R. (2005). Epistemological beliefs, mathematical problem­solving beliefs, and academic perform­ance of middle school students. The Elementary School Journal, 105(3), 289–304.

Schraw, G., Dunkle, M. E., and Bendixen, L. D. (1995). Cognitive processes in well­defined and ill­defined problem solving. Applied Cognitive Psychol-ogy, 9, 523–8.

Schraw, G. and Olafson, L. (2002). Teacher’s epistemological worldviews and educational practices. Issues in Education, 8(2), 99–148.

Schreiber, J. B. and Shinn, D. (2003). Epistemological beliefs of community college students and their learning processes. Community College Research and Practice, 27(8), 699–710.

Tsai, C.­C. (2002). Nested epistemologies: Science teachers’ beliefs of teach­ing, learning, and science. International Journal of Science Education, 24(8), 771–83.

Vosniadou, S. (2003). Exploring the relationships between conceptual change and intentional learning. In G. M. Sinatra and P. R. Pintrich (Eds.), Inten-tional conceptual change (377–406). Mahwah, NJ: Lawrence Erlbaum Associates.

A welcome and guide for the reader28

Vosniadou, S. and Brewer, W. F. (1992). Mental models of the earth: A study of conceptual change in childhood. Cognitive Psychology, 24, 535–85.

Vygotsky, L. S. (1978). Mind and society: The development of higher psychological processes. Cambridge, MA: Harvard University Press.

Walton, M. D. (2000). Say it’s a lie or I’ll punch you: Naive epistemology in classroom conflict episodes. Discourse Processes, 29(2), 113–36.

Wood, P. and Kardash, C. (2002). Critical elements in the design and analy­sis of studies of personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (231–60). Mahwah, NJ: Lawrence Erlbaum Associates.

Part II

Frameworks and conceptual issues

31

2 Manifestations of an epistemological belief system in preschool to grade twelve classrooms

Marlene Schommer­Aikins, Mary Bird, and Linda BakkenWichita State University

The purpose of this chapter is to present the potential implications of personal epistemology from a multidimensional perspective on pre­school to grade twelve research and instruction. The chapter is divided into six major sections. The first section provides a summary of the epistemological belief system as it was originally proposed by Schom­mer in 1990. The second section presents the more recent embedded systemic model of epistemological beliefs (Schommer­Aikins, 2004). The third and fourth sections set the ground work for others in developing more complex models. We do this by highlighting recursion and pos­sible scenarios of epistemological beliefs and cognitive development at specific points of recursion. Finally, we offer educational implications and research ideas that highlight epistemological beliefs and their place among other systems.

An epistemological belief system

The notion that personal epistemology should be conceptualized as an epistemological belief system was inspired by research that spans almost forty years. Perry (1968, 1970) conducted research with Harvard under­graduates that led him to conclude that undergraduates enter college thinking knowledge is simple, certain, and handed down by authority. By the time they reach their senior year, students’ thinking changes to believe knowledge is complex, tentative, and derived from reason and evidence. For the next twenty­two years other researchers (e.g., Kitch­ener and King, 1989; Baxter Magolda, 1987; Ryan, 1984), influenced by Perry, developed research agendas that explored varying aspects of personal epistemology. Those researchers and their colleagues tended to focus on a specific aspect of personal epistemology. While studying

Manifestations of an epistemological belief system 32

students across all grade levels, Kitchener and King (1989) focused on the justification of knowledge which involved issues of the source and certainty of knowledge. While studying high school students involved in mathematical problem­solving, Schoenfeld (1983) focused on students’ beliefs about omniscient authority, ability to learn, and speed of learn­ing. While studying young children and elementary students, Dweck and her colleagues (e.g., Dweck and Bempechat, 1983) focused on stu­dents’ beliefs about intelligence. While studying college students, Ryan (1984) focused on students’ beliefs about dualistic versus relativistic nature of knowledge. All of these researchers appreciated the impor­tance of students’ beliefs of the nature of knowledge and the nature of learning. How they differed is on which particular aspect of per­sonal epistemology was targeted for investigation. In 1990, Schommer suggested that in order to better understand personal epistemology, multiple beliefs need to be considered. She proposed that personal epistemology would be better conceptualized as a system of more­or­less independent beliefs. By system, she meant that there were multiple beliefs to be included. She initially hypothesized five beliefs that could take on a range of values. These hypothesized epistemological beliefs included beliefs about the stability of knowledge (certain, unchanging to tentative, evolving), structure of knowledge (simple, isolated bits to com­plex, integrated concepts), source of knowledge (omniscient authority to reason and evidence), speed of learning (quick or not­at­all to gradual), and ability to learn (fixed at birth to improvable).

It was hypothesized that as students matured and experienced envi­ronmental influences such as family, education, and culture, their beliefs matured, resulting in students believing that, more often than not, knowl­edge is complex, tentative, and derived from reason and evidence. And, more often than not, deep learning takes time and the ability to learn can improve. The most mature students know that these epistemic values will be context dependent. For example, in some situations knowledge and learning are simple and quick. But in many academic and everyday life situations knowledge is complex and learning is gradual.

By more­or­less independent, Schommer (1994) meant that it is pos­sible that the development of beliefs is asynchronous. For example, in a transitional point of development students may believe that knowl­edge is predominately complex. At the same time they may believe that knowledge is predominately certain. There are two important impli­cations from this assumption. First, teachers cannot assume that just because students are mature in one belief, they are mature in all of the epistemological beliefs. Second, by differentiating among beliefs, at any particular point in time, the theory predicts that some beliefs

Marlene Schommer­Aikins, Mary Bird, and Linda Bakken 33

may be affecting students, while other beliefs will not be relevant to the situation. For example, when assessing a student’s understanding of the tentativeness of a concept, one would predict that belief in certain knowledge would predict learning, whereas belief in the structure of knowledge may not be relevant for this particular learning event.

The inclusion of learning beliefs into personal epistemology has been controversial (Hofer and Pintrich, 1997). In 2004, Schommer­Aikins presented a model that separated the beliefs into those of learning and those of knowledge. Here we include both sets of beliefs in the episte­mological system for several reasons. For the sake of clarification, the epistemological belief system can be thought of as being composed of two subsets of beliefs. One subset is composed of three beliefs that are purely epistemological (knowledge stability, knowledge structure, and knowledge source). The other subset is composed of two beliefs that are epistemologically related (learning speed and learning ability).

There are three reasons why we continue to include the learning beliefs in this epistemological belief system. First, there is evidence that learning beliefs develop before knowledge beliefs (Schommer­Aikins et al., 2000). Second, we suspect that learning beliefs can either facili­tate or inhibit the development of knowledge beliefs. For example, if students have a strong belief in quick learning, they may develop a habit of careless studying. This would make it difficult for students to develop a belief in complex knowledge because they do not take the time to see the links between ideas. Third, beliefs about learning in conjunction with beliefs about knowledge often predict academic performance (e.g., Cano, 2005; Schoenfeld, 1983; Schommer et al., 1992; Qian and Pan, 2002). In other words, these two sets of beliefs are assumed to inti­mately interact with each other. Failure to include these beliefs when researching or teaching students makes our understanding of student cognition incomplete. Asserting that an epistemological belief system is composed of two subsets of beliefs – one purely epistemological and the other epistemologically related – may help alleviate confusion.

Epistemological beliefs are considered important because they are believed to influence student learning throughout all academic levels. Research provides supporting evidence for this hypothesis. For exam­ple, studies have linked epistemological beliefs to science learning (Bell and Linn, 2002; Elder, 2002), mathematical learning (De Corte et al., 2002; Schoenfeld, 1983; Schommer­Aikins et al., 2005), interpreta­tion of tentative information (Kardash and Scholes, 1996; Schommer, 1990), motivation (Bråten and Strømsø, 2004; Hofer, 1999; Schommer and Walker, 1997), and study strategy use (Schommer et al., 1992; Schrieber and Shinn, 2003).

Manifestations of an epistemological belief system 34

Epistemological inquiry is important because its effects are subtle. This subtlety is likely because epistemological belief effects are medi­ated by other variables (Schommer­Aikins, 2004). For example, path analyses indicate that the more students believe learning takes time, the more they believe math is useful. Subsequently, the more students believe math is useful, the better they are at solving mathematical prob­lems and overall academic performance. These findings were evident for middle school students (Schommer­Aikins et al., 2005) and college students (Schommer­Aikins and Duell, 2006). Cross­cultural research reports similar findings. Among Spanish high school students, the less students believed in quick learning and simple knowledge, the more likely they used deep processing approaches to learning. Subsequently, the more they used deep processing approaches to learning, the better they were at overall academic achievement (Cano, 2005).

Embedded systemic model

The challenge of studying epistemological beliefs goes beyond their subtlety due to mediated variables. What is important to consider is the role of epistemological beliefs to other systems and the effect of other systems on epistemological beliefs. In 2004, Schommer­Aikins pro­posed an embedded systemic model of epistemological beliefs. Simply put, epistemological beliefs do not function in a vacuum. She hypothesized that epistemological beliefs are enmeshed among other systems. And at any given moment, a thought or action is the result of the culminating effect of multiple systems.

This idea of embedded systems is similar to Bronfenbrenner’s (1979) proposed ecological systems theory of human development. More recently, Bronfenbrenner (Bronfenbrenner and Evans, 2000) suggested a bioecological systems model that includes process, person, context, and time. These models operate in a micro­ (family) and a meso­ (peer and school) system, as well as in macro­systems. Process, according to Bronfenbrenner, is actually a proximal process – an interaction among the systems that causes development to occur.

There are foreshadowings of the embedded systemic model that can be gleaned from existing research on epistemically related beliefs as well. Consider systems such as social systems, academic systems, family systems, and cognitive systems. Belenky et al. (1986) have uncovered the intimate ties between individuals’ ways of knowing and their epis­temological beliefs. Some individuals are strong believers in separate knowing. They approach new ideas by playing the devil’s advocate first. If they can be convinced about a new idea, then they are more open to

Marlene Schommer­Aikins, Mary Bird, and Linda Bakken 35

understanding. Others have a strong propensity to connected knowing. They approach new ideas with empathy. They take the perspective of another person, and then they think evaluatively. More recently Galotti et al. (1999) suggest that more cognitively mature individuals believe in both ways of knowing. Their challenge is to determine which way of knowing is needed depending upon the context. In a subsequent study (Schommer­Aikins and Easter, 2006), both connected knowing and separate knowing were shown to predict academic performance. But more importantly, this effect was mediated by belief in speed of learn­ing. In other words, the more students believed in connected knowing and separate knowing, the more they believed in gradual learning. Sub­sequently, the more students believed in gradual learning, the better they performed in a course that required reading, writing, and public presentations. In other words, students’ social system (connected and separate knowing) affects their epistemological belief (speed of learn­ing), which affects their academic performance.

Schommer­Aikins and Easter (2007) found evidence for the interac­tions among culture, study strategies, and epistemological beliefs. In a study comparing first generation Asian­Americans, beyond first gener­ation Asian­Americans, and Euro­Americans, academic performance was significantly different among these groups. Just as important were the significant differences found among groups in study strategies and epistemological beliefs. Asian­American groups had stronger beliefs in quick learning compared to Euro­American groups. This difference in beliefs about the speed of learning accounted for significant amounts of variance in group differences for study strategies and subsequent academic performance. In other words, culture was linked to episte­mological beliefs, study strategies, and academic performance. Future research may uncover more precisely the flow of influence, whether this flow of influence be unidirectional, reciprocal, and/or non­linear (Magnusson and Stattan, 1998).

Research and theory suggest that family systems influence episte­mological beliefs (Schommer, 1993). In a study comparing commu­nity college students with university students, examination of parental influence was found to be significant. The more parents encouraged their children to think independently, the less likely their children, as young adults, believed in simple knowledge. The more educated par­ents were, the less likely their children, as young adults, believed in quick learning.

In research examining the effects of epistemological beliefs on cog­nitive and behavioral strategies (Paulsen and Feldman, 2007), there is evidence of the combined effects of learning beliefs and knowledge

Manifestations of an epistemological belief system 36

beliefs. That is, each of these beliefs influences the other in a way that ultimately relates to students’ approaches to learning. For example, Paulsen and Feldman found that “naïve beliefs in [both] fixed abil­ity and in certain knowledge interact to intensify the negative effects of naïve epistemological beliefs on students’ use of deeper­processing strategies of elaboration and metacognition in their learning” (p. 384). And, although belief in quick learning or certain knowledge alone did not show a significant relationship to cognitive strategies, students with naïve beliefs in both quick learning and certain knowledge were more likely to use surface level rehearsal strategies. In short, the effects of a particular epistemological belief may not always be significant. Researchers may be missing the impact of this belief because they fail to look at the interaction and/or conditional effects of other epistemo­logical beliefs. For all intents and purposes, it is a failure to look at the links within and among systems that may result in missing important findings.

We hypothesize that cultural systems, ways of knowing, knowledge beliefs, learning beliefs, and classroom performance are all influenc­ing each other. Family systems, society as a whole, and physical envi­ronments can influence epistemological beliefs and each other. And epistemological beliefs can reciprocally influence the other systems (Bendixen and Rule, 2004).

In some senses, these ideas are intuitively obvious. Yet, being aware of such complexities is not the same as using these complexities in theory, research, or application. How do we formulate a theory that captures multiple systems? How do we test such an extensive theory? Schommer­Aikins (2004) proposed that researchers work in teams, coordinate their research questions, and use multiple methods of inquiry in an effort to develop dynamic­fluid models.

In order to help our readers begin to develop their own complex mod­els, we present a hypothetical example of embedded systems and how multiple systems could innervate epistemological belief development. We first focus on the notion of recursion. From this we generate stories about two hypothetical boys, one whose epistemological belief develop­ment is steady and one whose epistemological belief growth is limited. We reflect on how other systems contribute to the differences in their epistemological belief growth at points of recursion.

Recursion

Chandler et al. (2002) have offered a theory of recursion in which indi­viduals revisit epistemological development (i.e., absolutism, multiplism,

Marlene Schommer­Aikins, Mary Bird, and Linda Bakken 37

evaluativism) over and over again. We apply the theory of recursion specifically to epistemological beliefs; that is, epistemological beliefs are revisited and can be revised over different periods of an individual’s life­time (Schommer­Aikins, 2002). Revisiting does not guarantee revision; revisiting can result in perseveration at the same developmental level. Chandler et al. suggest two such time frames: “children between ages six and eight already appreciate that different persons can hold different beliefs about one and the same thing …” Approximately ten to twelve years later, “a period that is neatly bisected by an emerging capacity for abstraction (Eckensberger, 1983), or the onset of Piaget and Inhelder’s [sic: Inhelder and Piaget] (1958) formal operations” (Chandler et al., 2002, p. 162), these children now see ambiguity and uncertainty as per­vasive. Schommer­Aikins’ (2004) premise of the interrelationship and reciprocal interaction between epistemological beliefs about knowledge and epistemological beliefs about learning becomes part of this recur­sive cycle.

Building on Chandler et al.’s thinking, we would like to offer three time frames for recursion, that is, revising one’s epistemological beliefs. The first point would be at approximately five­ to seven­years­old when children have the initial cognitive ability to understand complexity (Steinberg and Meyer, 1995). For example, a six­year­old child, when asked if he was happy to attend a new event replied, “No, I’m scared.” “Why are you scared?” his grandmother asked. “I don’t know what’s going to happen there, if the teacher will be nice or if the children will like me.” The child’s remark suggests an awareness that future events can be uncertain and unpredictable. In other words, cognitively the child is able to use either inductive reasoning or is able to look at more than one variable simultaneously. During this particular time frame, however, if children are provided with sufficient “hands­on” activities, they are successful at recursion (Bakken et al., 2001). Successful recur­sion (revisiting leads to revision) also depends on successful inter­action with the multiple systems in individuals’ lives. Schommer­Aikins (2004) suggested that a feedback mechanism between systems (i.e., family, schools, peers) provides the means for epistemological beliefs to move beyond the simple to the more complex.

An example of this successful interaction comes from an intervention study. In this study of fifth­grade students, Bakken et al. (2001) found that all four sections of the fifth grade at a certain public school had stu­dents who were failing math and reading. Upon testing these students, the researchers found that the children were only capable of retaining one variable at one time. After instructional intervention, both classes who participated in the small groups with “hands­on” activities aimed

Manifestations of an epistemological belief system 38

at increasing their cognitive abilities were able to simultaneously con­sider several variables. Those children who did not participate in the small group activities did not improve. There were statistically signifi­cant differences between the two groups. In other words, the feedback mechanism between systems (evaluation, instruction, and classroom activity) resulted in students’ cognitive growth.

The second point occurs at about the beginning of adolescence or when the onset of formal operations occurs (Steinberg, 2007). At this point, early adolescents’ epistemological beliefs are revisited and modi­fied once again, although it may be situation­specific (or, as Chandler et al., 2002, would say, “retail”). In other words, epistemological beliefs will develop in one area but not in others. For example, evolution is taught as a fact by the instructor; however, some students study evolu­tion and question the gaps in the theory offered by the teacher. Here they are showing the capability of accepting ambiguity. It is then a sign that they no longer believe knowledge is a set of simple, certain facts. On the other hand, in another class in school, the students receive a “D” in a test. The students may accept this grade as a certain fact and interpret it to mean that they will not be able to improve their learning and subsequently decide whether they should bother to try to go to col­lege. Schommer­Aikins (2004) would suggest that the family system would have a powerful influence on the students’ decisions to continue their education. If the family system believes the ability to learn is fixed, then a poor grade is a sign of future failure (Holloway and Hess, 1985); this indeed could have an influence on recursion.

A third point of epistemological beliefs revision will occur in young adulthood if students experience a scholastic system that nurtures questioning and investigation rather than passive acceptance of facts. Steinberg (2007) suggests that, whereas most adolescents have the potential for using abstract thinking on a regular basis, it seems that environmental factors play a significant role in the complete acquisition of formal logic. Even with this caveat, it is not until young adulthood that individuals have the capability of epistemological beliefs that sup­port higher order thinking across domains (or “wholesale,” as Chandler et al., 2002, suggested).

We return to the notion of embedded systems: “there are mini­mally three basic influences on the learner’s life: family, peers, and teachers … Each of these groups could be considered a system, but more importantly, members of these groups have their own system of cultural views, ways­of­knowing beliefs, epistemological beliefs and, subsequently, expectations of classroom performance and self­ regulated learning” (Schommer­Aikins, 2004, p. 26). Hence, whether

Marlene Schommer­Aikins, Mary Bird, and Linda Bakken 39

students achieve epistemological belief growth during periods of recur­sion may be influenced by family, friends, and teachers, to name just a few potential influencers.

Scenarios of epistemological beliefs and cognitive development

Let us consider, next, epistemologically based instruction in the con­tent area of history using recursion. Let us also propose two possible paths of development, in which different bioecological systems interact (Bronfenbrenner and Evans, 2000). We will use examples of two boys, Robert and John. Robert comes from a stable environment. He lives with his mother, father, older brother, and younger sister. Both parents give their children a fair amount of attention. They are a middle income family. Both parents have a bachelor’s degree. He visits his grandpar­ents on a farm. The family also enjoys outings to the zoo, the county fair, and local museums.

John comes from an unstable environment. He lives in poverty with his aunt. They live in a small room in the middle of a big city. John does not have opportunities such as getting out of the big city, visiting the zoo, or going to museums. John does not have a concept of an extended family, since his aunt must work sixty hours a week in a minimum wage job in order to care for John, herself, and three other children. His aunt does not have time to devote to John. From time to time he walks through the neighborhood. Occasionally he will see fights, drug use, and other maladaptive behavior.

The contrasting home environments are meant to highlight the fam­ily and peer systems that may nurture or neglect the cognitive, emo­tional, social, and epistemological development of an individual. We conjecture that with a nurturing environment, the individual is likely to revisit his epistemological beliefs and grow accordingly. In a chaotic and/or impoverished environment, the individual is likely to remain unchanged or regress in his epistemological beliefs (Perry, 1968, 1970). We will describe the hypothetical progress of these two boys through the three time frames of recursion using explicit examples.

Time frame 1

Initial cognitive ability. The first point in our recursive scenario is when children are able to understand, simultaneously, differences in more than one variable (around the ages of five to seven years of age). Let us suppose that the first­grade teacher is ending a week­long unit in April

Manifestations of an epistemological belief system 40

on prairie animals. She is having the children prepare a diorama of prairie animal habitats. She provides the children with boxes, cutouts of animals, crayons, construction paper, scissors, and glue. Next, she provides step­by­step verbal instructions, followed by walking around the classroom and monitoring the children’s progress. The children work individually constructing the dioramas. John’s diorama has no background, no sunflowers, no clouds, no grass, one tree, one animal, and the animal (a prairie dog) is in a tree. John is finished in ten min­utes. John’s peers make fun of John’s animal in a tree; however John responds: “It’s right: my grandpa told me that prairie dogs live in a tree.”

Robert’s diorama has grass, several trees, three animals (a prairie dog, a snake, a frog), mounds for the prairie dog, and all animals are placed on the mounds. Robert takes two hours to complete his work. A girl peer tells Robert that the frog is incorrect; it should be a toad. Robert responds: “It should not; my grandpa told me it should be a frog.”

Here we see the sharp contrast between John and Robert. Episte­mologically, John’s work lacks complexity. His work seems rushed. His proof of factual knowledge is based on the authority of his grandfather. Epistemologically, Robert displays a modest degree of complexity with his more elaborate diorama. He takes his time to complete his project. However, he too relies on the authority of his grandfather to serve as proof for a contested frog in the scene.

How can the teacher facilitate epistemological development for these two boys? The teacher might question John about what he would see if he looked outside. As first graders are concrete thinkers, she may even take him to the window and ask him to name the various things that he sees in the outdoor environment. She may need to point out the grass and clouds as John may overlook the obvious. The teacher should take John back to revisit his diorama and ask him what he could do differ­ently to create a more outdoor setting. Another option the teacher could use is to place John’s diorama beside another diorama that is more complete and have him compare with her what he could include in his own diorama to make it more accurate. By utilizing these methods the teacher has incorporated comparing and contrasting, identification and categorizing (Markman, 1989). This allows John to see the inaccuracies and incompleteness of his schema (Gagné et al., 1993), which provides John with an opportunity to revise his beliefs about the source and structure of knowledge.

Robert may exhibit characteristics of being more advanced in under­standing the prairie habitat, but has misplaced the frog into the incor­rect environment. His interpretation may be due to confusion between

Marlene Schommer­Aikins, Mary Bird, and Linda Bakken 41

identifying characteristics of a frog for a toad, which could be incor­porated into a class lesson. She could ask, “How can we determine whether a frog or a toad should be in this picture?” The class would then be responsible for determining the method of problem­solving. The teacher has been sensitive to Robert by not saying that a frog would not be on the prairie, but has taken advantage of a learning opportu­nity. The teacher could simply place two columns on the board and the students could brainstorm attributes (Gagné et al., 1993) of both a toad and a frog, helping Robert determine which animal would be found on a prairie.

Interaction among systems. Developmentally, children at this age have the capability of beginning to understand complexity (Steinberg and Meyer, 1995). Thus, constructing the diorama is cognitively appro­priate. Peer criticism can provide the feedback that Schommer­Aikins (2004) suggested helps epistemological growth; it can also retard or stunt growth if the criticism is destructive. The academic system pro­vides another level in terms of context (Bronfenbrenner and Morris, 1998) in determining the proximal processes of development and sus­taining epistemological beliefs about knowledge and epistemological beliefs about learning. But what about the family system? Bronfen­brenner (Bronfenbrenner and Evans, 2000) suggests that the proximal process of development involves two possible outcomes: competence or dysfunction. And Hetherington and Stanley­Hagan (2002) indi­cate that, although families in transition place children at risk, there are many factors to consider before determining whether we will know which result will be John’s or Robert’s.

Time frame 2

Onset of formal operations. The second point in our recursive scenario is at the beginning of adolescence when the teenager begins to accept ambiguity along with the realization of abstraction. However, there are areas where the teen remains concrete in his thinking. Robert and John are now in seventh grade. It is again April and the history class has been studying the Revolutionary War. It is the end of the unit and the assign­ment is to construct a diorama of the Boston Tea Party. The teacher has allowed the class to self­select its groups. John’s and Robert’s tasks for their groups were to develop the lists that would be needed to construct the dioramas. Robert has included the following in his list: boats, barrels for tea (with labels indicating “tea”), confetti that represents loose tea, a large body of water where the boats are moored, people in period costumes. There is an argument within his group regarding the accuracy of his list. Robert

Manifestations of an epistemological belief system 42

responds: “I know it’s true. It says so in the history book. Besides, the Americans were heroes.” The diorama of his group reflects complexity and uses everything on Robert’s list in three­dimensional material.

John has included the following in his list: boat, tea bags, soldiers, water. There is an argument in his group about the tea bags. John responds: “It’s gotta be right because I saw it on TV. That’s how they make tea.” John’s group finishes quickly. The diorama is two­dimen­sional (the soldiers are paper cutouts).

Again, we see a contrast in development between the two boys. Epis­temologically, John continues to lack complexity. He still appears to believe learning is a quick process. He relies on information from popu­lar media (tea bags he saw on TV) that is far removed from the Bos­ton Tea scene to serve as the source and justification of his knowledge. Epistemologically, Robert continues to grow in the complexity of his thinking. For Robert, knowledge source and justification come from books written by authorities. What fails him is the accuracy of either his initial learning or the recall of the information provided in his his­tory book.

If the teacher is carefully observing the diorama preparations and the arguments occurring within the groups, what can he/she do? John’s group obviously has a limited background from which to draw for meaningful connections so the teacher could intervene with a sug­gestion. The teacher could tell John’s group that most people today do make tea from bags but in the late 1700s people made tea from brew­ing leaves. She could offer to bring loose tea and let the class taste the difference between loose and bagged tea (Bergin, 1999). While they drink the brewed tea, the class could watch a video on the Boston Tea Party and note the differences of people back then and today. At this point it would be advantageous to interject that television is for profit and the video is factual information for the development of under­standing an historical event accurately. The class could write down questions that they would have for the British who attended the Bos­ton Tea Party (Schraw et al., 2001). It is important to open the world­view for John as his experiences have limited his capacity to think critically and form necessary connections from previously acquired information. At this point the teacher could ask John’s group how they could construct a three­dimensional diorama, giving examples and possibly including the art teacher to help create British models out of clay or even showing the students how to make three­dimensional figures out of paper. This may encourage John to revise his belief in the structure of knowledge.

Marlene Schommer­Aikins, Mary Bird, and Linda Bakken 43

The teacher has an excellent opportunity, after watching the video about the Boston Tea Party, to make note to Robert that even though history books and videos depict actual events, there is no way that we can be sure that they are absolutely factual (Wineburg, 1991). Poetic license can be explained, showing how authors and directors may add or delete some factual elements to make a story more interesting or even to persuade the viewer. Factual documents were most prob­ably taken from a journal entry so it could be explained to Robert that a personal entry would most likely be subjective in contrast to objective.

By having the students work in groups, the teacher becomes a facili­tator to help the students find connections and show them ways to problem­solve together. John will grow epistemologically when he has supplementary information from which to draw, and this confidence in himself could be a benefit by gaining concrete experiences in the class­room (Schraw et al., 2001). By writing questions for the British, John will realize that he is part of a world larger than just the city he lives in and his life is formed from historical events and the people that were there. Robert will gain from the video in factual information just as John, but he will grow in the epistemological perspective that informa­tion from authority sources may be biased and does not always have to be accepted at face value.

Interaction among systems. Generally, young adolescents at this age reach brain maturation that allows them to begin to use the abstract thinking of hypothetical­deductive logic (Steinberg and Meyer, 1995). Thus, the three­dimensional diorama of the Boston Tea Party is an appropriate task. In terms of students self­selecting their peers to work with, Ormrod (2008) suggests that the teacher should select students to be in groups which will be productive. Again, peer criticism, if con­structive, can aid in developing more complex epistemological beliefs about knowledge and learning; if destructive, it can stunt those episte­mological beliefs. Academically, the context has provided evidence of the combined effects of learning beliefs and knowledge beliefs. That is, each of these beliefs influences the other in a way that ultimately relates to students’ approaches to effects of this belief. Last, let’s look at the family system. If we remember, Bronfenbrenner (Bronfenbrenner and Evans, 2000) suggested that the proximal processes of develop­ment could have two outcomes: competence or dysfunction. We see John and Robert at a later time in their progress toward adulthood. We do not know what is occurring in their family systems; however, according to research (Gottfried et al., 2003; Hetherington and Stanley­Hagen, 2002), maternal employment, divorce and single­parenthood

Manifestations of an epistemological belief system 44

by themselves do not provide the context for dysfunction. Therefore, we can come to no conclusions about Robert and John as yet.

Time frame 3

Formal operations. The third recursive revision occurs during young adulthood; however, it may also happen at the culmination of the high school experience. The young adult now fully accepts ambiguity and issues become interrelated. However, the individual can also distin­guish between the simple and the complex. He can see things in terms of shades of gray rather than just black and white (Wadsworth, 1978, 1996). We now meet Robert and John in April of their senior year of high school. They are in a United States Government class and are finishing a unit on the judicial system. The teacher has assigned the class to write a persuasive essay on lowering the drinking age to eighteen.

Robert has the following to say: These are my reasons for lowering the drinking age to 18: a person can be asked to fight for his country. A person can get married without parental consent. A person can get a driver’s license and pay taxes. However, there are some consequences for lowering the drink-ing age: teens are more likely to be the victims of driving accidents. There is research that even younger teens will start drinking. Drinking leads to irre-sponsible sex. I think drinking might affect teens’ academic performance. The scale seems to be about evenly balanced both ways; however, because most teens already drink, I think I would vote to lower the drinking age. I would add that there should be some responsibility added to the benefit. I’m not sure how it would be implemented, but it would be important. Perhaps there should be instruction in terms of how much and when to drink.

John has the following to say: These are my reasons for lowering the age to 18: we can go into the army. We can drive a car. Besides, everybody I know gets drunk. I can’t think of any reason not to lower the drinking age.

As the years have gone by, the epistemological belief chasm has grown between Robert and John. Epistemologically, Robert now embraces the breadth and depth of complexity. He sees more than one side to an issue. He relies on multiple sources of evidence. He is able to balance his final conclusion to allow younger people to drink, with a sense of responsibility instilled through education (Schommer­Aikins, 2004). Epistemologically, John remains almost unchanged from his earlier years. His argument is quick and simple. The voice of his message is self­absorbed with the constant reference to “we.”

What kind of feedback can a teacher give to help both students con­tinue to grow in their epistemological beliefs (Ames, 1990)? Robert

Marlene Schommer­Aikins, Mary Bird, and Linda Bakken 45

should be commended for demonstrating both sides of the issue and citing research. The teacher may use this opportunity to stress the importance of backing up a viewpoint with statistical and empirical documentation. An important question to ask Robert would be about his statement “since most teens already drink.” Is this statement founded on perceived information from personal experience or factually based? The teacher could ask Robert if he would rather adhere to laws that had been created based on fact or opinion? It is no surprise that John has based his case on his own opinion entirely. The teacher should contrast two persuasive arguments with one having a factual basis and the other from a personal basis so that John can see the difference. It should be explained to John that facts are more effective than using a personal viewpoint. John’s teacher could also help him see the line of reason­ing that would be used about lowering the drinking age from various perspectives. How would parents whose child has been killed by an intoxicated teen driver see his argument? What about a family who has a history of alcoholism; how would they view lowering the drinking age? The teacher could clarify that John will always have his own opinion, but when preparing a persuasive argument it is necessary to research the issue from all sides. It would be to John’s advantage if he could learn to see things from another’s viewpoint, making him more well­rounded and less self­focused.

Interaction among systems. Generally, brain maturation is complete. The late adolescent/young adult has the capability to use abstraction, hypothetical­deductive logic, and can accept ambiguity (Steinberg and Meyer, 1995). After studying the US judicial system, writing a persua­sive essay and presenting both sides of an argument should be possible for both John and Robert. The cultural system in their lives is likely one that accepts drinking by high school seniors. This is one of the contexts of their system. Academically, they have been challenged to provide arguments with evidence to substantiate their propositions. Schommer­Aikins (2004) suggests that cultural­relational views interact with individual epistemological beliefs about knowledge. This reciprocal interaction can increase the complexity of the epistemological belief system. However, if one’s cultural­relational views are to believe in simple, certain knowledge handed down by omniscient authority, the interactions may rather reinforce the individual’s epistemological belief in simple knowledge. Once again, family systems play an important role in the interaction of the proximal process: the contexts again are complex in that they are interrelated with those of the peers and the academic environment. However, the outcome of competence or dys­function becomes clearer.

Manifestations of an epistemological belief system 46

Summary

To summarize, a sharp contrast in epistemological belief develop­ment is shown between the two boys. In time frame 1, John is revisit­ing the source of knowledge and the structure. John keeps his beliefs that knowledge is based on authority and is factual in nature. In time frame 2, John is revisiting the structure of knowledge, the source of knowledge, and the speed of learning. Once again, John maintains that knowledge is based on authority, facts, and learning occurs quickly. In time frame 3, John is revisiting the same epistemological beliefs. Here we see slight change: John’s source of knowledge has changed from authority and facts to personal experiences. However, John still thinks knowledge is simple and quick to obtain.

Robert’s epistemological belief development shows revision at points of revisitation. In time frame 1, Robert is revisiting source of knowledge and structure of knowledge. Whereas Robert maintains authority as the source of knowledge, he has revised his belief in the structure of knowl­edge to include a degree of complexity. In time frame 2, Robert revisits the structure, source, and justification of knowledge. Now source of knowledge includes text as well as people and empirical evidence. He continues to grow in his awareness of the complexity of knowledge. In time frame 3, Robert revisits the structure and source of knowledge. Robert expects knowledge to be justified from multiple sources of evi­dence that should converge. He is now able to sense between highly complex knowledge structures and simple ideas among the complex.

In terms of recursion, we see both John and Robert revisiting earlier development levels: specifically, their beliefs in the source, structure, justification, and certainty of knowledge. What we see are two dif­ferent trajectories. John’s epistemological belief growth is slow and limited. In other words, he perseverates rather than revises. On the other hand, Robert’s revisitations indicate substantial growth in the source, structure, and justification of knowledge through these three time frames.

Educational implications

When reflecting on the implications of epistemological beliefs in edu­cation, it is useful to think of several systems all interacting simultane­ously. These systems include social interactions, cultural expectations, and historical backgrounds. The classroom can be seen as a source of social interaction. Furthermore, any specific class develops its own norms and expectations within a short amount of time. In other words,

Marlene Schommer­Aikins, Mary Bird, and Linda Bakken 47

it develops a culture. Finally, each individual, including teachers and students, bring with them their own personal and academic history. Vygotsky’s (1978; cited in Miller, 2002) theory of socio­cultural devel­opment contributes to our understanding of this interaction among these systems.

According to Vygotsky, children’s minds are inherently social. In other words, it is the socio­cultural­historical time frame that defines children’s cognitive development. Rather than looking within an indi­vidual for causes of behavior and learning, Vygotsky would suggest that it is more meaningful to address the socio­cultural contexts. He indi­cated that learning occurs first on an intermental (or social) level in terms of relationships and interactions with others. It then becomes intramental (or individual) learning as a result of these relationships and interactions when children internalize what they have learned from others (Wink and Putney, 2002). From Vygotsky’s perspective, learn­ing is dynamic and has a dialectical relationship with children’s socio­cultural worlds (Vygotsky, 1986).

Wink and Putney (2002) suggest that, in this context, classrooms become part of a very complex puzzle. Based on the Vygotskian premise, the teacher, the family, and the student “act in dialogue with each other, jointly constructing what counts as knowledge” (p. 63). During these interactions, students bring with them their historical experiences which have helped shape them thus far in their development. Wink and Putney also indicate that the classroom culture is socially constructed by the interactions between teacher and students. What students and teachers have learned from their pasts contribute to their participation in the present.

In this case, the history that we allude to is individuals’ epistemologi­cal belief systems. Beliefs that could influence students in the classroom could be their own beliefs, the beliefs of their parents, and the beliefs held by the teacher. The teacher’s own epistemological beliefs (which are usually subconsciously held) may influence student performance in the classroom (Calderhead, 1996). If the teacher holds the belief that learning must be quick to take place or it will not take place at all, it would impede students from getting more practice or extended time to work on a skill because the teacher would stop once a concept was covered. Another way that a teacher’s beliefs could affect the students is if he/she believed that all information should be provided to students by one omniscient authority. This would then limit the classroom to the lecture­and­test format which is not conducive to learning for students who have varying learning styles such as kinesthetic or visual (Gardner, 1993).

Manifestations of an epistemological belief system 48

Teacher education can play an important role in encouraging an epis­temologically based classroom. Teachers would be trained at the under­graduate educational level in how their own beliefs shape their students. DiPietro (2004) suggests that teachers are most flexible to accept sug­gestions in the preservice years as students. Some veteran teachers who have deeply ingrained ideologies that the teacher is the all­knowing fact provider and students are passive learners may benefit from inservice training (Richardson, 1996).

In general terms, what would the epistemologically based classroom look like? In order to encourage students to embrace the complexity and evolution of knowledge, the teacher pushes his/her students to think deeply. For example, he/she encourages students to look for connec­tions among concepts within text, with their prior knowledge, and with concepts found in the world beyond themselves. Some class activities allow students to discover multiple solutions; some class activities may result in no definite solutions at all. Rather, students generate possi­bilities that are context specific. Ideally these types of activities occur across the curriculum which encourages both domain­specific and domain­general epistemological belief development.

These epistemologically driven activities may be very challenging for some students. Effective teachers would encourage students to not give up if they have not grasped a concept immediately. An explanation would be provided by teachers telling the students that developing a level of expertise takes time and experience. With patience, tenacity, and hard work they will approach mastery of a subject.

Would an epistemologically based classroom look different at the dif­ferent grade levels? The answer to this is yes and no. The teacher would function much as a motivator and facilitator in all of the classrooms to varying degrees. In the lower grades the teacher would take a larger role in guiding students to discover and construct knowledge. In higher grades students may assume a greater role in the discovery, construc­tion, and application of knowledge.

The ability of students to grasp abstract thinking, thus affecting critical and higher level thinking skills, varies by grade level as well. Teachers can ask students why questions. Why did an event happen? Why is there inequality in the world? Why can’t we find the answer to certain health dilemmas? Teachers can ask these questions at all grade levels. But they will anticipate different types and degrees of sophistication in students’ responses. This questioning of why pro­vides students with an opportunity to see that education is not just a series of facts, but all of the information is integrated and critically examined. Although students may not be developmentally ready

Marlene Schommer­Aikins, Mary Bird, and Linda Bakken 49

to generate lofty answers, asking them why is likely to encourage epistemological belief development.

Implications for future research

If we can agree that epistemological beliefs are best understood amid the ebbs and flows of other systems, how can we possibly investigate them? Does this seem to be impossible?

We assert that is it difficult and cumbersome. However, we do not see it as impossible. We offer four major suggestions. The first two sug­gestions come from Schommmer­Aikins’ (2004) original presentation of an embedded systemic model. The third suggestion applies ways of knowing theory (Belenky et al., 1997) to researchers’ collaborative as well as competitive spirits. The fourth suggestion focuses on the nature of hypotheses. We then offer several examples of hypotheses in the con­text of our two hypothetical boys.

First, researchers and practitioners can work together to develop aspects of a systemic model. Each individual or group can make a con­tribution to an overall model. And, obviously, competing embedded systemic models would provide a great opportunity to capture a model (or models) that help explain learning and lead to better suggestions for instruction.

Second, researchers need to coordinate their work. It would be over­whelming for only one person, or only one group, to completely test a complex embedded model. By carefully laying out specific hypoth­eses from a model and conducting research in a coordinated fashion, researchers can share results and build the next set of hypotheses to be tested. For example, some aspects of the model may need to be tested simultaneously. Other aspects of the model may need to be tested in sequence, one result leading to the next question and that result leading to still another question.

Third, research teams should embrace diversity in method and measurement. By this we mean that individual researchers or research teams will have strength in certain methods and/or measurements. When results come in from researchers who excel in their methodol­ogy/measurement, interpretations might be different than the norm. For example, some researchers could use interviews, others could use questionnaires, others could use observations, and others could use spatial­visual representations. A range of interpretations needs to be considered when results are conflicting. Consider the possibilities: each measure is capturing a unique nuance of the phenomenon; some meas­ures are more sensitive than others. The conceptualization behind the

Manifestations of an epistemological belief system 50

measure is creating the conflict. A new measure can be derived based on the input of the conflicting measures. In other words, the interpreta­tion is less like a horse race that concludes: who is the winner? Rather, it is more like a shared experience of exploring the students and their worlds that concludes: what can we learn from this? In some sense, this reflects the use of both connected knowing and separate knowing in the interpretation of the results (Belenky et al., 1997).

Fourth, in order to get a sense of hypotheses that can be generated to test an embedded systemic model, we suggest that the research ques­tions (or hypotheses) reflect multiple systems. Here are three examples for you to ponder.

Example 1. If epistemological beliefs do not mature at a reasonable pace, will other cognitive abilities be inhibited in growth, misused, or maladaptive? For example, will holding an excessively strong belief that knowledge does not change, result in a failure to adapt to changes in the world around them, failure to listen to other people’s views, and failure to change in moral, social, and academic domains?

Example 2. In order to make a change in epistemological beliefs, does a culmination of multiple systems need to occur or can a change in one system be enough? For example, will teachers’ supportive epistemo­logical instruction be enough for epistemological growth or will the support of other systems, such as peers, family, and environmental cir­cumstances, be necessary for change?

Example 3. What happens when there is an epistemological belief tug­of­war? For example, if a student’s own developing epistemological beliefs differ from his teacher’s, and his teacher’s beliefs differ from his parents, who will the child model? Or will the child switch beliefs depending on her/his environment? Or will the child reject either lead, and attempt to find his own epistemological path? And will this deci­sion depend upon the age/cognitive maturity of the child (Bendixen and Rule, 2004; Bronfenbrenner and Evans, 2000; Schommer­Aikins, 2004)?

We leave you with one final thought. As you attempt to create models and/or test models, keep the everyday world in mind. Do not get lost in the fog of lofty language. Do not let the elegant sounds of pseudo­philosophical hyperbole let your mind get clouded as to basic goals of research. Although there are many goals to research, we like to believe that understanding human thinking, developing optimal instruction, and providing critical intervention for students who are at the brink of academic despair will help us keep our theory, our research, and our results in the field of epistemological beliefs important to students, teachers, parents, and policy­makers.

Marlene Schommer­Aikins, Mary Bird, and Linda Bakken 51

R EF ER ENCES

Ames, C. (1990). Motivation: What teachers need to know. Teachers College Record, 91, 409–21.

Bakken, L., Thompson, J., Clark, F. L., Johnson, N., and Dwyer, K. (2001). Making conservationists and classifiers of preoperational fifth­grade chil­dren. The Journal of Educational Research, 95(1), 56–61.

Baxter Magolda, M. B. (1987). Comparing open­ended interviews and stand­ardized measures of intellectual development. Journal of College Student Personnel, 28(5), 443–8.

Belenky, M. F., Clinchy, B. M., Goldberger, N. R., and Tarule, J. M. (1997). Women’s ways of knowing. New York: Basic Books.

Bell, P. and Linn, M. C. (2002). Beliefs about science: How does science instruction contribute? In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (321–46). Mahwah, NJ: Lawrence Erlbaum Associates.

Bendixen, L. D. and Rule, D. C. (2004). An integrative approach to personal epistemology: A guiding model. Educational Psychologist, 39, 69–80.

Bergin, D. A. (1999). Influences on classroom interest. Educational Psycholo-gist, 34, 87–98.

Bråten, I. and Strømsø, H. I. (2004). Epistemological beliefs and implicit the­ories of intelligence as predictors of achievement goals. Contemporary Edu-cational Psychology, 29(4), 371–88.

Bronfenbrenner, U. (1979). The ecology of human development. Cambridge, MA: Harvard University Press.

Bronfenbrenner, U. and Evans, G. W. (2000). Developmental science in the 21st century: Emerging theoretical models, research designs, and empirical findings. Social Development, 9, 115–25.

Bronfenbrenner, U. and Morris, P. A. (1998). The ecology of developmental processes. In R. M. Lerner (Ed.), Handbook of child psychology, Vol. 1: The-oretical models of human development (5th edn., 535–84). New York: Wiley.

Calderhead, J. (1996). Teachers: Beliefs and knowledge. In D. C. Berliner and R. C. Calfee (Eds.), Handbook of educational psychology (709–25). New York: MacMillan.

Cano, F. (2005). Epistemological beliefs and approaches to learning: Their change through secondary school and their influence on academic per­formance. British Journal of Educational Psychology, 75, 203–21.

Chandler, M. J., Hallett, D., and Sokol, B. W. (2002). Competing claims about competing knowledge claims. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

De Corte, E., Op ’t Eynde, P., and Verschaffel, L. (2002). “Knowing what to believe”: The relevance of students’ mathematical beliefs for mathe­matics education. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing (297–320). Mahwah, NJ: Lawrence Erlbaum Associates.

DiPietro, K. (2004). The effects of a constructivist intervention on pre­service teachers. Educational Technology and Society, 7(1), 63–77.

Manifestations of an epistemological belief system 52

Dweck, C. S. and Bempechat, J. (1983). Children’s theories of intelligence: Con­sequences for learning. In S. G. Paris, G. M. Olson, and H. W. Stevenson (Eds.), Learning and Motivation in the Classroom (239–56). Hillsdale, NJ: Lawrence Erlbaum Associates.

Eckensberger, L. H. (1983). Research on moral development. German Journal of Psychology, 7, 1095–244.

Elder, A. D. (2002). Characterizing fifth grade students’ epistemological beliefs in science. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

Gagné, E. D., Yekovich, C. W., and Yekovich, F. R. (1993). The cognitive psych-ology of school learning (2nd edn.). New York: Longman.

Galotti, K. M., Clinchy, M. B., Ainsworth, K. H., Lavin, B., and Mansfield, A. F. (1999). A new way of assessing ways of knowing: The attitudes toward thinking and learning survey (ATTLS). Sex Roles, 40(9/10), 745–66.

Gardner, H. (1993). Frames of mind: The theory of multiple intelligences (10th edn). New York: Basic Books.

Gottfried, A. E., Gottfried, A. W., and Bathurst, K. (2003). Maternal and dual­earner employment status and parenting. In M. H. Bornstein (Ed.), Handbook of parenting, Vol. 2: Biology and ecology of parenting (2nd edn., 207–29). Mahwah, NJ: Lawrence Erlbaum Associates.

Hetherington, E. M. and Stanley­Hagen, M. (2002). Parenting in divorced and remarried families. In M. H. Bornstein (Ed.), Handbook of parenting, Vol. 3 (2nd edn., 287–315). Mahwah, NJ: Lawrence Erlbaum Associates.

Hofer, B. (1999). Instructional context in the college mathematics class­room: Epistemological beliefs and student motivation. The Journal of Staff, Program, and Organizational Development, 16(2), 73–82.

Holloway, S. D. and Hess, R. D. (1985). Mothers’ and teachers’ attribution about children’s mathematics performance. In I. E. Sigel (Ed.), Parental belief systems: The psychological consequences for children (177–99). Hillsdale, NJ: Lawrence Erlbaum Associates.

Hofer, B. and Pintrich, P. (1997). The development of epistemological theor­ies: Beliefs about knowledge and knowing and their relation to learning. Review of Educational Research, 67(1), 88–140.

Inhelder, B. and Piaget, J. (1958). The growth of logical thinking from childhood to adolescence. New York: Basic Books.

Kardash, C. M. and Scholes, R. J. (1996). Effects of preexisting beliefs, episte­mological beliefs, and need for cognition on interpretation of controversial issues. Journal of Educational Psychology, 88, 260–71.

Kitchener, K. S. and King, P. M. (1989). The reflective judgement model: Ten years of research. In M. L. Commons, C. Armon, L. Kohlberg, F. A. Richards, T. A. Grotzer, and J. D. Sinnot (Eds.), Adult development 2: Mod-els and methods in the study of adolescent and adult thought (63–78). New York: Praeger.

Magnusson, D. and Stattin, H. (1998). Person­context interaction theories. In R. M. Lerner (Ed.), Handbook of child psychology, Vol. 1: Theoretical models of human development (5th edn., 685–759). New York: Wiley.

Marlene Schommer­Aikins, Mary Bird, and Linda Bakken 53

Markman, E. M. (1989). Categorization and naming in children. Cambridge, MA: MIT Press.

Miller, P. H. (2002). Vygotsky and the sociocultural approach: Theories of develop-mental psychology (4th edn., 367–419). New York: Worth Publishers.

Ormrod, J. E. (2008). Educational psychology (6th edn.). Upper Saddle Hill, NJ: Prentice Hall.

Paulsen, M. B. and Feldman, K. A. (2007). The conditional and interaction effects of epistemological beliefs on the self­regulated learning of college students: Cognitive and behavioral strategies. Research in Higher Educa-tion, 48(3), 353–401.

Perry, W. G. Jr. (1968). Patterns of development in thought and values of students in a liberal arts college: A validation of a scheme (ERIC Document Reproduc­tion Service No. ED 024315). Cambridge, MA: Bureau of Study Counsel, Harvard University Press.

(1970). Forms of intellectual and ethical development in the college years: A scheme. Troy, MO: Holt, Rinehart and Winston.

Qian, G. and Pan, J. (2002). A comparison of epistemological beliefs and learn­ing from science texts between American and Chinese high school stu­dents. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (365–85). Mahwah, NJ: Lawrence Erlbaum Associates.

Richardson, V. (1996). The role of attitudes and beliefs in learning to teach. In J. Sikula (Ed.), The handbook of research in teacher education (2nd edn., 102–19). New York: MacMillan.

Ryan, M. P. (1984). Monitoring text comprehension: Individual differences in epistemological standards. Journal of Educational Psychology, 76, 248–58.

Schoenfeld, A. H. (1983). Beyond the purely cognitive: Belief systems, social cognitions, and metacognitions as driving forces in intellectual perform­ance. Cognitive Science, 7(4), 329–63.

Schommer, M. (1990). The effects of beliefs about the nature of knowledge on comprehension. Journal of Educational Psychology, 82, 498–504.

(1993). Comparisons of beliefs about the nature of knowledge and learn­ing among post secondary students. Research in Higher Education, 34, 355–70.

(1994). Synthesizing epistemological research: Tentative understand­ings and provocative confusions. Educational Psychology Review, 6, 293–319.

Schommer, M., Crouse, A., and Rhodes, N. (1992). Epistemological beliefs and mathematical text comprehension: Believing it is simple does not make it so. Journal of Educational Psychology, 84(4), 435–43.

Schommer, M. and Walker, K. (1997). Epistemological beliefs and valuing school: Considerations for college admissions and retention. Research in Higher Education, 38, 173–86.

Schommer­Aikins, M. (2002). An evolving theoretical framework for an episte­mological belief system. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (103–18). Mahwah, NJ: Lawrence Erlbaum Associates.

Manifestations of an epistemological belief system 54

(2004). Explaining the epistemological belief system: Introducing the embedded systemic model and coordinated research approach. Educa-tional Psychologist, 39, 19–29.

Schommer­Aikins, M. and Duell, O. K. (2006). General and mathematical epis-temological beliefs. Paper presented at the American Educational Research Association, San Francisco.

Schommer­Aikins, M., Duell, O. K., and Hutter, R. (2005). Epistemological beliefs, mathematical problem­solving, and academic performance of middle school students. The Elementary School Journal, 105(3), 289–304.

Schommer­Aikins, M. and Easter, M. (2006). Ways of knowing and epistemo­logical beliefs combined effect on academic performance. Educational Psychologist, 26(3), 411–23.

(2007). Epistemological beliefs’ contributions to study strategies of Asian-Americans and Euro-Americans. Paper presented at the American Educational Research Association, San Francisco.

Schommer­Aikins, M., Mau, W., Brookhart, S., and Hutter, R. (2000). Under­standing middle students’ beliefs about knowledge and learning using a multidimensional paradigm. The Journal of Educational Research, 94(2), 120–27.

Schraw, G., Flowerday, T., and Lehman, S. (2001). Increasing situational interest in the classroom. Educational Psychology Review, 13, 211–24.

Schreiber, J. B. and Shinn, D. (2003). Epistemological beliefs of community college students and their learning processes. Community College Research and Practice, 27(8), 699–710.

Steinberg, L. (2007). Adolescence (8th edn.). Boston: McGraw­Hill.Steinberg, L. and Meyer, R. (1995). Childhood. New York: McGraw­Hill.Vygotsky, L. S. (1986). Thought and language (revised and edited by A. Kozulin).

Cambridge, MA: MIT Press.Wadsworth, B. J. (1978). Piaget for the classroom teacher. New York: Longman. (1996). Piaget’s theory of cognitive and affective development. New York:

Longman.Wineburg, S. S. (1991). On the reading of historical texts: Notes on the breach

between school and academy. American Educational Research Journal, 28(3), 495–519.

Wink, J. and Putney, L. (2002). A vision of Vygotsky. Boston: Allyn and Bacon.

55

3 Epistemic climate in elementary classrooms

Florian C. Feucht University of Toledo

In today’s elementary classrooms, children are expected to acquire knowledge and skills to ensure that they will be able to maintain and further the cultural, economic, and scientific accomplishments of our societies as responsible citizens. As a result, early on in their school career, elementary school students are exposed to a diversity of knowl­edge and skills explained to them by their teachers, systematized in curricula and syllabi, and conveyed through instruction, school books, and other educational materials.

The goal of this chapter is to consider how the actual nature of the knowledge and skills to be learned is portrayed and perceived in elementary classrooms. That is, how is the nature of knowledge and knowing presented by teachers, instruction, and educational materials? From an educational perspective, is knowledge conveyed as black­and­white, stripped from its complexity and ambiguity, and assessed in a right­or­wrong fashion? From a more disciplinary perspective, are dif­ferent ways of knowing along with their characteristic subjectivities and tentativeness acknowledged by teachers, instruction, and educational materials? Correspondingly, what are the beliefs of elementary school students about knowledge and knowing and are their epistemic beliefs influenced by (such) common classroom components? Do elementary students believe in the objectivity of knowledge and the existence of absolute truth? Or do they think of knowledge as being more delicate and subject to change? Taking all of these different classroom com­ponents and influences into consideration, the overarching questions of this chapter include: what is the epistemic climate of elementary classrooms, what does it look like, and what is its possible educational potential?

A total lack of empirical insight on the nature of epistemic climates – not only with regard to elementary classrooms – is apparent throughout varied fields of educational interest and research. Only little is known about elementary teachers’ epistemic beliefs, and even less regarding elementary school students. This is comparable to what is (not) known

Epistemic climate in elementary classrooms56

about the epistemological influence of instruction and educational mate­rials on elementary students, specifically, and among all components involved, in general. Therefore, the goal of this chapter is (1) to accu­mulate what has been written – empirically and theoretically – about the nature of epistemic climates, and its components and relations, and (2) to assimilate what is known in one overarching conceptual frame­work of epistemic climate.

To achieve this goal, the chapter is laid out in the following man­ner: in the first section, the reader’s attention is brought to the more implied awareness of epistemic climates in the field of personal epis­temology research before different steps are taken to define a working definition of epistemic climates and to rationalize the educational model of personal epistemology as a legitimate conceptual framework for the purpose of this chapter. In the second section, existing empirical and theoretical literature on the epistemic climate, and its components and relations, is reviewed. Within the context of elementary education, the review focuses on students’ personal epistemology, teachers’ personal epistemology, and the epistemological underpinnings of instruction and knowledge representations, such as curricula and school books. It concludes with important issues stemming from the integration of these five reviewed components. The final section discusses implica­tions for future research and educational practices, specifically in the light of context­specific, school­subject­specific, and developmental aspects of epistemic climates. A brief summary concludes the overall chapter.

Conceptualizing epistemic climate

Several scholars in the field of personal epistemology have pointed out that a person’s beliefs about knowledge and knowing may be influenced by different factors situated within the person’s immediate and/or dis­tant environment. In this line of thinking, Vygotsky’s (1978) work on socio­cultural development is often referenced to explain how a soci­ety’s prevalent culture might impact the epistemic beliefs of individuals and their development through cultural artifacts, language, and other processes of enculturation (e.g., Bendixen and Rule, 2004; Feucht and Bendixen, in press). Furthermore, Bronfenbrenner’s (1979) ecological system model is applied to systematize the proximity of the influential factors in the personal environment (e.g., Schommer­Aikins, 2004; Palmer and Marra, 2008). Furthermore, combinations of these socio­constructivist (Vygotsky, 1978) and systemic (Bronfenbrenner, 1979) frameworks are used in combination to describe epistemological factors

Florian C. Feucht 57

and processes in a person’s environment as well as culture (e.g., Haerle and Bendixen, 2008). Overall, there is the understanding that the epis­temologies of individuals are influenced (at least to a certain extent) by a variety of surrounding factors and processes, which might be epistemic or epistemological in their nature themselves. This epistemic influence is considered an important aspect of epistemic climates (Bendixen and Rule, 2004; Feucht, 2008; Haerle and Bendixen, 2008).

In the field of personal epistemology, many theoretical assumptions and research objectives acknowledge implicitly the existence of epi­stemic climates. For example, from a cross­cultural perspective, the conduct of comparative research and related issues are based on the assumption that cultural contexts of different societies may influence the epistemic development of their citizens differently (e.g., Chan and Elliott, 2004; Khine, 2008; Maggioni et al., 2006; Quian and Pan, 2002). From an educational perspective, Perry (1970) stressed the importance of enduring learning environments on the maturation of personal epistemologies in general, while Hammer and Elby (2002) assume that students’ personal epistemologies are activated through their immediate context from situation to situation. From a concep­tual perspective, part of Hofer and Pintrich’s (1997, 2002) definition comprises the context sensitivity of epistemological theories, while Schommer­Aikins (2004) remodeled her epistemological belief system on the basis of system theories. Despite the understanding that epis­temic characteristics within a person’s context and culture influence his/her personal epistemology, only few publications explicitly address the epistemic climate as a construct (e.g., Bendixen and Rule, 2004; Feucht, 2008; Haerle and Bendixen, 2008).

A working definition of epistemic climate

In a very broad understanding, epistemic climate can be defined as a context encompassing different epistemic factors (e.g., math problems and news commentary) and processes (e.g., problem­solving and school education) that interact and influence a person’s epistemology. Such epistemic factors and processes are situated within the micro, meso, and macro levels of a person’s environment (Bronfenbrenner, 1979; e.g., individual person, school, and society) and are reflective of arti­facts and processes of enculturation (Vygotsky, 1978; e.g., books, peers, and language acquisition). From birth to death, a person is constantly exposed to different epistemic climates, which differ from context to context (e.g., home, school, church, and work). Therefore, a context

Epistemic climate in elementary classrooms58

completely free of epistemic factors and processes, an epistemic vac­uum, is practically impossible.

Bendixen and Rule (2004) were the first to address the nature of epistemic climates explicitly. In contrast to the above broad definition, they focus in their definition on the reciprocal influence of epistemic beliefs among different people in an immediate (micro level) environ­ment (i.e., reciprocal causation). Specifically, they posit that a student’s epistemic beliefs can be influenced and advanced in its development through the epistemic beliefs of their peers, parents, and teachers. Fur­thermore, Bendixen and Rule explain that epistemic climates can occur within and outside the classroom context and can have an important influence on society­wide changes in epistemic beliefs. The latter assumption broadens their conceptualization or at least its influence to the societal (macro) level.

Later, the construct of epistemic climate was re­conceptualized by integrating more components relevant to classroom education in pre­school through tertiary education (Feucht, 2008; Haerle and Bendixen, 2008). This definition of epistemic climate is grounded in the educa-tional model of personal epistemology (Haerle, 2006; Figure 3.1) and used as a working definition in this chapter:

The term epistemic climate defines the nature of knowledge and knowing of a classroom emerging from the personal epistemologies of students and their teachers, as well as from the epistemological underpinnings of instruction and knowledge representations along with the reciprocal relations among these four components. An epistemic climate is sensitive to its context and school subject, and influential on epistemic development.

Considering the current discussion on the terminology used in the field of personal epistemology (e.g., Bendixen, 2002; Kitchener, 2002; Murphy et al., 2007), the term epistemic climate seems to be more suit­able as a label for this construct than epistemological climate. In the light of other existing terms, such as personal epistemology, the term classroom epistemology can be used as a synonym for epistemic climate.

The educational model of personal epistemology

The educational model of personal epistemology (EMPE; Figure 3.1) is the product of a theoretical contemplation of why the personal episte­mologies of ninety­eight elementary students I interviewed were simi­lar within a classroom (sample) and different as compared to other classrooms (Haerle, 2006). That is, the initial purpose of the model was (1) to explain how typical components of education (i.e., teachers,

Florian C. Feucht 59

instruction, educational materials) influence the personal epistemol­ogy of learners and (2) to speculate how these components could be used – from an educational standpoint – to inform the development of instruction to enhance the personal epistemology of students within their own learning environment. More recently, the model has also been applied (3) to function as a conceptual framework theorizing epistemic climate and (4) as a research framework to operationalize classroom and cross­cultural research on personal epistemology (Feucht, 2008; Haerle and Bendixen, 2008). The purpose of this section is to introduce the model and to provide a more detailed account of its foundations and synthesis.

Within a classroom context, the EMPE focuses on epistemic char­acteristics of students, their teacher, the content knowledge taught by the teacher and learned by the students, and the instruction used to convey this knowledge. More specifically, the components of the model comprise (1) learners’ personal epistemology (i.e., the beliefs of learners about knowledge and knowing), (2) the teacher’s personal epistemology (i.e., the beliefs of the teacher about knowledge and knowing), (3) epis­temic knowledge representations (i.e., the epistemic messages that are embedded in content knowledge, such as in school curricula and school books), and (4) epistemic instruction (i.e., the epistemic messages that are embedded in the teaching methods used to teach content knowl­edge). Finally, (5) the reciprocal relations, depicted with double­headed arrows, are the glue of the model. They describe the interdependency of the components and define the dynamic nature of the model.

Epistemic instructions

Epistemic knowledgerepresentations

Learners' personalepistemology

Teacher's personalepistemology

Figure 3.1: The educational model of personal epistemology

Epistemic climate in elementary classrooms60

The foundation and synthesis of the educational model of personal epistemology

Four theoretical frameworks can be identified, which explicitly address epistemological aspects of education. Two of these frameworks are rooted in the field of curriculum and instruction and describe epis­temological elements important to constructivist lesson planning and conceptual change learning (Kattmann et al., 1996; Westphal, 1990), while the other two frameworks discuss the role of personal epistem­ology for learning and instruction from the educational psychology perspective (Bendixen and Rule, 2004; Hofer, 2001). All frameworks encompass two or three components existing in the working definition of epistemic climate and the EMPE (see Table 3.1).

The model of life-problem-centered pedagogy. In the field of curriculum and instruction, Westphal (1990) developed a framework to guide teach­ers in the construction of lesson plans that revolve around everyday life problems. The model follows a constructivist approach proposing that it is essential to assess and account for students’ prior knowledge in the development of instruction and lesson planning. It is assumed that tailoring the instructional approach accordingly would better facilitate the acquisition of the content knowledge to be learned and the recon­struction of existing knowledge, respectively. More specifically, in light of Westphal’s life­problem­centered pedagogy, he suggests to screen students’ prior knowledge and the content knowledge as well for six cognitive structures: self­realization, heteronomy, epistemology acquisi-tion, self­shaping, sustainability, and development. He argues that these structures, which are essential to human behavior and development and identifiable as such in all accounts of human knowledge, can be used to strategically guide the development of instruction and lesson plan­ning as they are naturally conducive to human learning. The structure of epistemology acquisition is of special interest here. Westphal (1990) describes the structure as the cognitive need to conceive complexity and to systematize the environment and self. His conceptualization of epistemology acquisition clearly overlaps with Hofer and Pintrich’s (1997) epistemic dimension of simplicity of knowledge addressing the simple, complex, and connected nature of knowledge (see Haerle, 2006, for a detailed comparison). In summary, to inform classroom educa­tion Westphal suggests not only to assess students’ prior knowledge and content knowledge regarding their epistemological underpinnings, but also to explicitly feature epistemological structures as possible layouts of the actual instruction and lesson plans. His model of life-problem-centered pedagogy is a framework accountable for the development of instruction

Tab

le 3

.1. O

verv

iew

of m

odel

s de

scri

bing

cha

ract

eris

tics

of e

pist

emic

clim

ate

Ed

uca

tion

al m

odel

of

per

son

al e

pis

tem

olog

y S

ynth

esis

of

curr

icu

lum

an

d in

stru

ctio

n an

d ed

uca

tion

al p

sych

olog

y m

odel

s

Lea

rner

sT

each

erIn

stru

ctio

nK

now

led

geR

elat

ion

s

Hae

rle

(200

6)L

earn

ers’

per

sona

l ep

iste

mol

ogy

Tea

cher

’s

pers

onal

ep

iste

mol

ogy

Epi

stem

ic in

stru

ctio

nE

pist

emic

kn

owle

dge

repr

esen

tati

ons

Rec

ipro

cal r

elat

ions

Cu

rric

ulu

m a

nd

in

stru

ctio

n m

odel

s

Wes

tpha

l (19

90)

Stu

dent

s’

unde

rsta

ndin

g of

th

e co

mpl

exit

y an

d sy

stem

atiz

atio

n of

the

con

tent

kn

owle

dge

to b

e le

arne

d

–In

stru

ctio

n th

at

help

s st

uden

ts t

o un

ders

tand

the

co

mpl

exit

y an

d sy

stem

atiz

atio

n of

the

co

nten

t kn

owle

dge

The

co

mpl

exit

y an

d sy

stem

atiz

atio

n of

the

con

tent

kn

owle

dge

to

be le

arne

d

Stu

dent

s’ u

nder

stan

ding

of

the

com

plex

ity

and

syst

emat

izat

ion

of t

he

cont

ent

know

ledg

e an

d th

e co

mpl

exit

y an

d sy

stem

atiz

atio

n of

the

con

tent

kn

owle

dge

itse

lf a

re a

naly

zed

to in

form

th

e co

nstr

ucti

on o

f in

stru

ctio

n (i

.e.,

one­

dire

ctio

nal i

nflue

nces

)

Kat

tman

n et

al.

(199

6)E

pist

emol

ogic

al

belie

fs a

s an

asp

ect

of

stud

ents

’ eve

ryda

y co

ncep

tion

–In

stru

ctio

n th

at h

elps

st

uden

ts t

o ac

quir

e do

mai

n ­sp

ecifi

c ep

iste

mol

ogy

of t

he

scie

ntifi

c co

ncep

t to

be

lear

ned

Epi

stem

olog

y of

the

sci

enti

fic

conc

ept

and

its

rela

ted

dom

ain

Stu

dent

s’ e

pist

emic

bel

iefs

and

the

ep

iste

mol

ogy

of t

he s

cien

tific

con

cept

are

co

mpa

red

and

cont

rast

ed t

o in

form

the

co

nstr

ucti

on o

f in

stru

ctio

n st

ep b

y st

ep

(i.e

., re

cipr

ocal

influ

ence

s am

ong

all t

hree

co

mpo

nent

s)

Ed

uca

tion

al p

sych

olog

y m

odel

s

Hof

er (

2001

)S

tude

nts’

ep

iste

mol

ogic

al

theo

ries

Tea

cher

’s

epis

tem

olog

ical

th

eori

es

Cla

ssro

om t

asks

and

pe

dago

gica

l pra

ctic

es–

Tea

cher

’s e

pist

emol

ogic

al t

heor

ies

influ

ence

the

ir c

hoic

e of

cla

ssro

om t

asks

an

d pe

dago

gica

l pra

ctic

es, w

hich

the

n in

fluen

ce s

tude

nts’

epi

stem

olog

ical

th

eori

es (

i.e.,

one­

dire

ctio

nal i

nflue

nce)

Ben

dixe

n an

d R

ule

(200

4)

Stu

dent

s’ p

erso

nal

epis

tem

olog

y T

each

er’s

pe

rson

al

epis

tem

olog

y

– –

Stu

dent

s’ a

nd t

each

er’s

per

sona

l ep

iste

mol

ogie

s ar

e in

fluen

ced

by f

eedb

ack

loop

s (i

.e.,

reci

proc

al in

fluen

ce)

Epistemic climate in elementary classrooms62

and lesson planning across all school subjects. Due to its focus on actual content knowledge and students’ prior knowledge, it also accounts for domain­ and context­specific ways of knowing effectively.

The model of educational reconstruction. Kattman and colleagues ( Kattman et al., 1996; Duit et al., 2005) developed the model of educational reconstruction as a guide for teachers and other educational professionals in the development of domain­/subject­specific instruction, lesson plans, curricula, and school books, to name a few. The model originated in the field of curriculum and instruction, more specifically in science educa­tion, and is currently applied across many different school subject areas. The model encompasses three components that are fundamental for the development of instruction: (1) scientific clarification, (2) analysis of students’ everyday conceptions, and (3) construction of instruction. For the purpose of instructional development, Kattmann and his colleagues suggest to screen scientific and educational materials to clarify its most appropriate version/representation and to analyze students’ everyday conceptions about the concept to be learned. Part of this screening is also to clarify the epistemological aspects of the scientific concept and to assess students’ personal epistemologies embedded in their everyday conception. In a constant process of comparing and contrasting, these two sources of information (including their epistemic aspects) are used to develop new instruction that allows students to re­construct their every­day conceptions towards a more scientific conception. In other words, Kattmann and his colleagues consider epistemological characteristics of students and content knowledge in the development of instruction and lesson plans. For example, students’ beliefs that knowledge in biology is objective and certain might be challenged by introducing them to com­peting and more complex knowledge claims as an aspect of a biology lesson. The model of educational reconstruction is innovative in its purpose to provide domain­specific instruction and lesson plans for scientific concepts, while putting students’ everyday conceptions at the heart of their development.

The working model of how epistemological theories influence classroom learning. From an educational psychology perspective, Hofer (2001) briefly hypothesizes a working model to describe how epistemological theories might influence learning in a classroom context. Three of the model’s factors are directly linked to epistemic climate and, therefore, are of interest here: (1) teachers’ epistemological theories, (2) classroom tasks and pedagogical practices, and (3) students’ epistemological theo­ries. Hofer posits that teachers are influenced by their epistemologi­cal theories when they choose their classroom tasks and pedagogical practices. These tasks and practices are perceived and interpreted by

Florian C. Feucht 63

students through their epistemological lenses and influence their epis­temological theories. Hofer assumes that when students change their epistemological theories, their perception and interpretation of class­room instruction will alter accordingly. Possible reasons for such an epistemic change might be the influence of different types and methods of instruction on students’ personal epistemologies. Details about the actual epistemic change process are not discussed. In summary, Hofer (2001) assumes in her model a one­directional influence of teachers’ epistemological theories on students’ epistemological theories through their choice of classroom tasks and pedagogical practices.

The integrative model of personal epistemology. Bendixen and Rule (2004) developed a theoretical model to explain the processes of epistemic belief change by integrating theoretical and empirical work in the field of edu­cational psychology. The integrative model of personal epistemology focuses on internal processes of epistemic change (e.g., epistemic doubt), but also addresses intrapersonal factors, which might impact the internal change process (i.e., peers, teachers, parents, and society). Of interest here is what Bendixen and Rule describe as epistemic climate inside and outside the classroom context. That is, the reciprocal influence of students’ personal epistemology on the personal epistemology of their peers, teachers, and parents. In other words, epistemological differences in an individual’s environment (i.e., epistemic beliefs of others) may trigger a mechanism of change. Bendixen and Rule ascribe a positive effect to these reciprocal relationships. They propose that feedback loops within a group of peo­ple could stimulate and multiply advanced epistemic beliefs. This effect is referred to as reciprocal causation and may play an important role in society­wide changes of personal epistemology.

Comparing and contrasting the models. The conceptualization of EMPE is informed by all four models, their components, and interrela­tions (see Table 3.1). Each model contributes different aspects to the model’s foundation. In the light of classroom epistemology, the model of life-centered pedagogy (Westphal, 1990) and the model of educational reconstruction (Kattman et al., 1996), both anchored in the field of cur­riculum and instruction, account for epistemic characteristics of stu­dents (i.e., learners’ personal epistemology), content knowledge (i.e., epistemic knowledge representations), and instruction (i.e., epistemic instruction). More specifically, in the process of developing instruction and lesson plans, both models propose to account for students’ personal epistemology and epistemological underpinnings of the content knowl­edge to be learned. That is, instruction should be explicitly reflective of epistemic assumptions identified in students’ prior knowledge and content knowledge. One difference between both models is that

Epistemic climate in elementary classrooms64

Westphal’s model describes a one­directional influence of epistemolog­ical aspects of students and content knowledge to inform the develop­ment of instruction and lesson plans, while Kattmann and colleagues’ model proposes a reciprocal relationship among all of its components (i.e., scientific clarification; students’ everyday conceptions; construc­tion of instruction) and their epistemic characteristics. The second difference is evident in the models’ conceptualization of epistemic char­acteristics. Westphal proposes epistemology acquisition as a cognitive structure that focuses exclusively on simple, complex, and/or connected aspects of knowledge and knowing, while Kattmann and colleagues refer generally to domain­specific, epistemological aspects of students and scientific concepts without providing a more detailed categorization. Both models differ to the models in the field of educational psychology (Bendixen and Rule, 2004; Hofer, 2001) in their account of epistemic characteristics of content knowledge (i.e., knowledge representations) and the absence of teachers as influential epistemic components (i.e., teachers’ personal epistemology) in classroom epistemology.

Hofer’s (2001) working model of how epistemological theories influence classroom learning and Bendixen and Rule’s (2004) integrative model of personal epistemology address epistemic characteristics from a more psy­chological perspective. Both models describe influential factors on per­sonal epistemology within the classroom context and the relation among them. Bendixen and Rule discuss the epistemic beliefs of students and peers (i.e., learners’ personal epistemology) and the epistemic beliefs of teachers (i.e., teacher’s personal epistemology), while Hofer also men­tions classroom tasks and pedagogical practices and their epistemic influence (i.e., epistemic instruction) as additional components. Apart from the number of components relevant to conceptualize epistemic climate, the models differ in the nature of the relations among compo­nents. Hofer assumes that teachers’ epistemological theories influence their instruction and pedagogy implicitly, which in turn influences their students’ epistemological theories. Bendixen and Rule, in con­trast, emphasize a reciprocal relation between students’ and teachers’ epistemic beliefs. Furthermore, while Hofer takes a neutral stance on the broader influences among these components, Bendixen and Rule envision a great potential of advancing personal epistemology inside and outside of the classroom. Synthesizing these four models from the curriculum and instruction and educational psychology literature into one model, the EMPE, is an innovative and promising approach ensur­ing its relevance for the overall field of education.

Synthesizing the educational model of personal epistemology. The compar­ing and contrasting of these models illustrates five components that are

Florian C. Feucht 65

vital in describing epistemic climate (see Table 3.1). The first four com­ponents of the EMPE are identifiable across all models: (1) learners’ personal epistemology (e.g., part of students’ everyday conceptions or epistemic beliefs/epistemological theories), (2) teacher’s personal epis­temology (e.g., epistemic beliefs/epistemological theories), (3) epistemic instruction (e.g., part of instruction and pedagogical practices), and (4) epistemic knowledge representations (e.g., part of content knowledge or scientific concepts). Furthermore, all four models describe or imply different influences among their components, either as one­directional or reciprocal in nature. In the EMPE, these relationships are described as (5) reciprocal relations, which describe the interdependency of the components and define the nature of the model as a dynamic system. Overall, six reciprocal relations are conceptualized in the EMPE. None of the reviewed models addresses all of these relations and/or their reciprocal nature. In other words, some relations and their reciprocal nature were hypothetically added to complete the conceptualization of the model.

The EMPE was purposefully designed to form the three­ dimensional shape of a pyramid. Within the classroom context, this presentation seems to be most appropriate to capture the outreaching nature of its epistemic components into each and every classroom corner and to depict their reciprocal relations accordingly. Because the model incor­porates insights gained from different educational disciplines, it has great potential to inform research, theory, and practice within and across a variety of educational fields. For example, it can be utilized as a framework to operationalize research on epistemic climate (e.g., as a whole, selected components, and/or relations) and to guide theoretical discussions and provide educational implications. In the following sec­tion, it is used to structure the literature review on epistemic climates of elementary classrooms.

A literature review on epistemic climate in elementary classrooms

Within the scope of this chapter the goal for this section is to review literature on the construct of epistemic climate and its components and processes. The review portrays predominately research and theory of epistemic characteristics within the context of elementary education. The identified literature stems from a variety of educational fields, such as educational psychology, curriculum and instruction, teacher educa­tion, and sociology. No literature could be identified that focuses on the construct of epistemic climate per se; at least beyond the little literature

Epistemic climate in elementary classrooms66

already reviewed in the earlier section addressing the working defini­tion of epistemic climate (i.e., Bendixen and Rule, 2004; Feucht, 2008; Haerle and Bendixen, 2008). However, a wide range of publications can be identified that focus on selections of different factors and relations evident in the working definition of epistemic climate. For this reason, the EMPE (see Figure 3.1) is applied as a theoretical framework to guide and structure the review at hand.

The review contains the following sections: (1) personal epistemology of teachers, (2) personal epistemology of learners, (3) epistemic instruc­tion, and (4) epistemic knowledge representations. The last section, (5) reciprocal relations and emerging issues, integrates the four previ­ous sections and addresses, therein, the reciprocal relations of epistemic climate. The review captures the essence of a more detailed review (see Feucht, 2008) and acknowledges the literature already discussed previ­ously in this chapter. Essentially, the purpose of the review is not only to portray the potential nature of epistemic climates in elementary class­room settings, but also to further reinforce the conceptualization and synthesis of the EMPE.

Personal epistemology of teachers

Only a very small amount of publications exist that address the personal epistemology of elementary teachers. For this reason – as an exception to the other sections – the literature is broadened to elementary and secondary school teachers, most of which are published in the field of educational psychology, teacher education, and higher education. They encompass two theoretical pieces (Patrick and Pintrich, 2001), seven exploratory studies (Brownlee, 2001; Chan and Elliot, 2000; Johnston et al., 2001; Schraw and Olafson, 2002; Sinatra and Kardash, 2004; Tsai, 2002; White, 2000), and three intervention studies (Brownlee et al., 2001; Howard et al., 2000; Gill et al., 2004).

Developmental levels. The existing research demonstrates that the per­sonal epistemology of teachers ranges on a developmental continuum, which can be roughly divided into three different categories (e.g., abso­lutist, multiplisit, and evaluativist; Kuhn et al., 2000). Absolutist teach­ers perceive teaching as transferring knowledge from teachers as experts to students as naïve learners. Their learning objectives for students are to acquire objective content knowledge in a passive manner. This has been labeled as objective learning models (Howard et al., 2000), mon­ological discourse patterns (Johnston et al., 2001), realist worldviews (Schraw and Olafson, 2002), traditional teacher beliefs (Tsai, 2002), and departing absolutists (White, 2000). In contrast, multiplist teachers

Florian C. Feucht 67

facilitate learning environments in which students actively construct their own personal understanding of the content knowledge. Expert views on content knowledge, curricula, and school books are often disapproved of because they perceive knowledge as more subjective and tentative. This has been labeled as constructivist learning models (Howard et al., 2000), dialogical discourse patterns (Johnston et al., 2001), relativist worldviews (Schraw and Olafson, 2002), constructiv­ist teacher beliefs (Tsai, 2002), and intuitive relativists and aspects of selective relativists (White, 2000). Finally, evaluativist teachers pro­mote learning activities in which students collaboratively construct knowledge on the basis of a shared understanding and are asked to commit to their understanding. Because these teachers perceive con­tent knowledge as context­dependent and tentative, they complement curricula with multiple, additional knowledge sources. This has been labeled as contextualist worldviews (Schraw and Olafson, 2002), reflec­tive relativists, and aspects of informed relativist (White, 2000).

Inconsistencies and change. Epistemic change in teachers has been described as gradual (Brownlee et al., 2001; Tsai, 2002; White, 2000). That is, teachers’ personal epistemology progresses step­by­step along different epistemic dimensions (e.g., certainty and structure of knowl­edge), which can cause epistemic inconsistencies. Such internal incon­sistencies are often found in preservice and novice teachers. These inconsistencies may decrease with the increment of teaching experi­ences (i.e., teachers’ personal epistemologies become more consistent over the years; Patrick and Pintrich, 2001; Tsai, 2002).

The few existing intervention studies show that preservice and inservice teachers’ personal epistemologies can be actively changed through different instructional approaches. Two studies exposed teachers to constructivist instruction, such as argument activation, refutational text, and conceptual change learning strategies (Gill et al., 2004; Howard et al., 2000). In a third study, teachers were asked to write reflective statements on their personal epistemology as part of a year­long educational psychology course (Brownlee et al., 2001). All three studies found that such instructional interventions can be suc­cessful in advancing teachers’ personal epistemology, even in a con­siderably short amount of time (Gill et al., 2004). In addition, teachers with more advanced (relativist) personal epistemologies seemed to be more receptive to epistemic interventions (Gill et al., 2004), which is in line with the discussion that teachers with more naïve beliefs about knowledge and knowing are more resistant to educational reform and new classroom practices (e.g., Patrick and Pintrich, 2001; Sinatra and Kardash, 2004).

Epistemic climate in elementary classrooms68

Impact on classroom practices and students. Several studies provide evi­dence that teachers’ personal epistemologies (see previous absolutist, multipilist, and evaluativist teacher descriptions) have an impact on the epistemic climate of their classrooms. That is, teachers’ personal episte­mologies influence their perception of content knowledge (e.g., single cur­riculum/textbook versus multiple knowledge sources), their preferences regarding instructional approaches (e.g., teacher­centered versus student­centered), and their understanding of the student as a learner (e.g., passive recipient versus active constructor) (Howard et al., 2000; Johnston et al., 2001; Schraw and Olafson, 2002; Tsai, 2002; White, 2000).

Johnston and colleagues’ (2000) case study, which was based on a six­month observation and a series of interviews, illustrates how the per­sonal epistemology of two elementary school teachers impacts not only their educational practice in English, but also their students’ personal epistemology. The teacher who employed dialogical patterns of class­room discourse viewed knowledge as complex, constructed, and highly related to individual experiences. She viewed her role as supporting stu­dents in becoming better and independent thinkers and appreciated the complexity of knowledge by promoting greater student participation, multiple perspectives, and student­centered tasks and assessments. The teacher who followed monological discourse patterns valued a sin­gle truth and perceived her role as an authority who delivers facts and corrects errors. She preferred the discussion of non­controversial issues and viewed differences in students’ knowledge as errors rather than as individual interpretations. Students who were exposed to the dialogical patterns of classroom discourse perceived literacy as a meaning­making activity for which their own experiences and those of their peers were very important. They expected to take part in shared knowledge pro­duction and appreciated their own and others’ knowledge differences in this process. In contrast, students in the classrooms in which monologi­cal patterns were practiced emphasized clear concepts of technical and performance success and perceived themselves as passive consumers of knowledge. This contrasting approach of teachers’ discourse practice, pedagogical beliefs, and student perceptions enabled Johnston and col­leagues to identify how these teachers differ in their personal episte­mologies. This approach, furthermore, illustrated that “epistemologies can be traced from teacher to student through the discursive practices of the classroom” (Johnston et al., 2001, p. 230).

Summary. The personal epistemologies of teachers can be divided into different developmental levels (e.g., absolutist, multiplisit, and evaluativist) and are subject to change. They become internally more consistent with increasing teaching experience and can be advanced

Florian C. Feucht 69

with interventions. Teachers with more advanced epistemic beliefs are more receptive to epistemic development and less resistant to educa­tional reform. Furthermore, teachers’ epistemic beliefs impact their educational classroom practice (e.g., instruction, content knowledge, and educational materials) and the epistemic beliefs of their students.

Personal epistemology of learners

Little theoretical and empirical literature exists that focuses on the personal epistemology of elementary school students. Two theoretical (Chandler et al., 2002; Walton, 2000) and nine empirical publications can be identified with this particular focus. Five of these studies are of an exploratory nature and are examined in this section (Elder, 2002; Haerle, 2006; Kuhn et al., 2000; Mansfield and Clinchy, 2002). The remaining four are instructional and/or intervention studies and, there­fore, reviewed in the next section on epistemic instruction. All pub­lications were located in the fields of educational and developmental psychology.

Theoretical assumptions about epistemic development. Two very different developmental assumptions exist in the field of personal epistemology and these might be due to an overall lack of research on children’s per­sonal epistemology. The first theoretical assumption refers to a late-on-set and linear development of personal epistemologies (Chandler et al., 2002). For example, King and Kitchener (1994) and Perry (1970), who researched mainly college­aged students, assumed that children in ele­mentary and secondary schools would have naïve (i.e., dualistic) beliefs about knowledge and knowing that would start to further develop in late adolescence and mainly in combination with exposure to higher educa­tion. This assumption has been challenged by Chandler and colleagues, who argue for an early-onset and spiral-like development of personal epis­temology (e.g., Boyes and Chandler, 1992; Chandler et al., 2002). They propose that elementary school children can already hold sophisticated beliefs and progress repeatedly though different epistemological lev­els until a certain epistemological maturity is gained. In line with this proposition, it has been argued that classroom education and teacher behavior could suppress epistemic development in learners (Boyes and Chandler, 1992; Chandler et al., 2002; Walton, 2000). In more recent research, the assumption of early­onset development, recursion, and suppression have been used as explanations for the diversity of personal epistemology identified in elementary school students.

Empirical findings on epistemic development. Burr and Hofer (2002) seem to have identified the commencement of epistemic beliefs (i.e., egocentric

Epistemic climate in elementary classrooms70

subjectivity) in preschool children applying “false­belief” tasks from the field of theory of mind. Leaving unexplored the age range of four to five years, four exploratory studies on elementary school students investigate children at the age of nine years and older and are discussed in the following section (Elder, 2002; Haerle, 2006; Kuhn et al., 2000; Mansfield and Clinchy, 2002). These studies reveal that elementary school students hold a diversity of personal epistemologies that range on a developmental continuum from naïve to sophisticated beliefs.

Kuhn and colleagues (2000) investigate the development of episte­mological understanding in third­, fifth­, eighth­grade, and older stu­dents. They developed a fifteen­item instrument to assess the degree to which knowledge is perceived as objective, subjective, or both. The study revealed three stages. The first stage, absolutist, is characterized by the belief that knowledge is objective. The multiplist stage follows which is, in contrast, described as a subjective understanding of knowl­edge. The final stage, evaluativist, is described as the reintegration of objectivity into subjectivity. Their results support that development from absolutist to multiplist occurs in a systematic order. The find­ings of this cross­sectional study are of particular interest because the evidence that third­, fifth­, and eighth­graders can already develop an evaluativist understanding of knowledge is supportive of the early onset assumption of epistemic development (Chandler et al., 2002).

Mansfield and Clinchy (2002), also interested in epistemic develop­ment, chose to conduct a longitudinal interview study with children at the ages of ten, thirteen, and sixteen. Similar to Kuhn and colleagues (2000), they were interested in the integration of objectivity and sub­jectivity as an indicator of epistemological development. The older stu­dents became, the more they became aware of the complexity of both objective and internal knowledge. Similarly, they began to perceive inconsistencies in their prior knowledge, to question the certainty of authority, and to distrust their sensory perception as a reliable resource for knowledge acquisition. Most of the students in the sample had a subjective understanding of knowledge, which is comparable to the muliplist level (Kuhn et al., 2000), while only one student ranked on the highest level, integrating objective and subjective knowledge. This is in contrast to Kuhn and colleagues’ study in which considerably more children were identified to be on the evaluativistic level.

While the previous two studies’ goals were to tap epistemic devel­opment, the following studies focus on the exploration of epistemic beliefs within the same age group. Haerle (2006) conducted a large­scale qualitative study with German fourth­graders. The interviewed children verbalized diverse beliefs about the origin, acquisition, and

Florian C. Feucht 71

verification of knowledge. They viewed knowledge as (1) a human invention, (2) a result of random trial and error, (3) a result of goal­directed investigations, (4) apparent through sensory perception, (5) a biological inheritance, (6) given by God, (7) derived from personal experience, and/or (8) derived from logical thought. While most of the fourth­graders believed that, historically, the origin of knowledge would be the human mind, they referred to sensory perception and second­hand knowledge when they spoke of their own everyday learning. Fur­thermore, this study showed that children can have epistemic beliefs similar to adults (i.e., absolutist, multiplist, and evaluativist), except that their “units” of knowledge were less abstract and smaller in scope compared to those typically described in the literature on adult per­sonal epistemology.

In contrast to the more domain­general approach of Haerle’s study, Elder (2002) investigated the epistemological beliefs of fifth­graders in science specifically. The analysis of an administered questionnaire with open questions and Likert­scaled items revealed a range of naïve and sophisticated beliefs about the epistemology of science. This result is similar to Haerle’s (2006) and Kuhn and colleagues’ (2000) studies. In addition, most of the students described science as an engagement in activities, lacking the understanding that science involves the effort to explain phenomena. Elder assumed this understanding of science to be possibly caused by a dominance of hands­on activities typical to these students’ science education. In other words, the epistemic underpin­nings of the instruction used influenced the students’ beliefs about the nature of science.

Summary. The reviewed studies demonstrate that elementary school­aged children have a diversity of epistemic beliefs. Ranging on a develop­mental continuum of absolutistic and evaluativist beliefs, most children appeared to be multiplists, the intermediate level. These results support the early­onset theory of epistemic development (e.g., Chandler et al., 2002) and challenge the late onset theory (King and Kitchener, 1994; Perry, 1970). Possible explanations for the developmental diversity of personal epistemology in elementary students might be their recursive development and/or their suppression and encouragement through classroom practices.

Epistemic instruction

Again, only a small amount of literature can be identified that addresses epistemological underpinnings of instruction in the context of elemen­tary education. Three theoretical (Hofer, 2001; Scheffler, 1965; Wade,

Epistemic climate in elementary classrooms72

1975) and five empirical (Boscolo and Mason, 2001; Louca et al., 2004; Johnston et al., 2001; Smith et al., 2000; Steinbring 1991) publications can be identified in the fields of educational psychology, curriculum and instruction, and subject pedagogy.

Theoretical assumptions. In the previously reviewed working model on how epistemological theories influence classroom learning, Hofer (2001) briefly assumes a one­directional line of causation between instruction and the epistemological theories of learners. In particular, Hofer sug­gests that the nature of the chosen types and methods of instruction might be influential on students’ personal epistemology. Unfortunately, no further details on this aspect are provided. While Hofer takes a more neutral stance on this issue, Scheffler (1965) and Wade (1975) out­line positive and negative aspects of epistemological underpinnings of instruction.

Wade (1975) discusses how instruction can influence the personal epistemology of learners by making the case that the epistemological underpinnings of teaching models and instruction should always match the epistemology of the subject knowledge taught. That is, epistemo­logical inconsistencies between these two could cause an inappropriate and conflicting epistemic understanding of the subject knowledge in the learner. As an example, Wade (1975) explores the teaching of creation­ism through the application of a Piagetian teaching model. The epis­temic conflict here is that knowledge about God’s creation of Earth and humankind is grounded in faith, while the Piagetian teaching model is justified through scientific reasoning. Wade argues that the Piagetian teaching model, which promotes the instruction of analytical knowl­edge in accordance with the learner’s cognitive development, fosters beliefs that knowing in religious studies is an analytical affair as well. This epistemic influence can be avoided by selecting a teaching model that is in accordance with the epistemology of the subject knowledge. To Wade, ignoring the impact of epistemic instruction is educationally irresponsible. For example, possible epistemic inconsistencies or mixed messages can confuse students and/or alter their personal epistemolo­gies in an unwanted way.

Scheffler (1965) has a different perspective on epistemic instruction. Rather than advocating an epistemological consistency between instruc­tion and subject knolwedge, he proposes three different but equal models of epistemic instruction. The aim of the impression model is to transmit simple knowledge pieces into the learner’s mind through sen­sory experiences and language. Because the teacher in this model is an active provider of knowledge and students are passive recipients, teach­ers have a large impact on the shaping of students’ minds. In the insight

Florian C. Feucht 73

model, knowledge is understood as a matter of insight. Language is used not to impress knowledge pieces into the learners’ mind; but rather it is instrumentalized to engage learners actively in their own search for reality and insight. Following the rule model, students should learn how to assess new knowledge and insights by values and reasoning eminent in science, morality, and culture. In other words, character develop­ment and independent critical thinking is to Scheffler an important part of teaching.

Empirical research. The existing research on the influence of instruc­tion does not follow such complex frameworks. The reviewed studies focus either on the exploration of different epistemic instruction or aim to measure its influence on students’ personal epistemology following an experimental design. While Johnston and colleagues (2001), reviewed earlier, were able to provide evidence that students’ personal epistemol­ogy might be influenced by their teachers’ instructional approaches, Louca and colleagues (2004) and Steinbring (1991) explored different instructional approaches to foster learning in general and to stimulate learners’ personal epistemologies in particular. The first study describes how the use of epistemic metaphors can help third­grade students to distinguish between ontological and mechanistic forms of knowledge in a science unit on color changes in leaves, while the second study examines how the mathematical concepts of probability and change can be taught to fifth­grade students using a socio­constructivist teach­ing approach.

The experimental designs of Boscolo and Mason (2001) and Smith and colleagues (2000) provide evidence that personal epistemologies of fifth­graders in history and sixth­graders in science can be advanced. In Boscolo and Mason’s study, students worked on a geographical unit of work on the discovery of America, in which conflicting historical documents were discussed as an instructional activity. The experimen­tal group contrasted from the control group as students used individual writing to express, reflect, and monitor their process of knowledge­building. The data analysis showed that students in the experimental group had a better conceptual understanding of the content knowledge and their personal epistemologies were further advanced.

Smith and colleagues (2000) investigated the influence of a construc­tivist and a traditional teaching approach on students’ personal episte­mology. In the constructivist condition (experimental group), students discussed their hypotheses and conceptions as a community of learn­ers, while in the traditional condition (control group) students were engaged in problem­solving and critical thinking. The data analysis showed that students who were in the constructivist intervention gained

Epistemic climate in elementary classrooms74

the understanding that knowledge is not simple and absolute and scien­tific conceptions derive from a process of criteria­guided evaluation.

Summary. Instruction can influence students’ personal epistemology due to their epistemic messages. Research has shown that construc­tivist instruction, in particular, is conducive to advancing epistemic beliefs. From a theoretical perspective, it is advisable that inconsisten­cies between instruction and content knowledge should be avoided by the teacher while students need to learn to gain insight and evaluate knowledge as independent thinkers.

Epistemic knowledge representations

A large amount of literature exists that addresses epistemological mes­sages evident within knowledge representations, such as curricula, newspapers, and the world wide web. However, when focusing the search at the elementary level, and on educational knowledge represen­tations such as curricula, work sheets, textbooks, board writings, and other educational media, the literature is very limited. Interestingly, all nine of the reviewed publications focus on the epistemic underpinnings of curricula and are published in the field of curriculum and instruc­tion. Seven, more philosophical, pieces elaborate on curricula as reflec­tions of scientific knowledge structures or worldviews, while two articles present research on the epistemic impact of curricula on teachers. In this section, the term curricular epistemology or epistemologies is used to refer to the constructions of epistemological underpinnings or mes­sages of curricula.

Curricular epistemology as knowledge structures. Researchers and schol­ars, such as Pines (1982), Golin (1997), and Dobson and Dobson (1987) refer to curricular epistemology as knowledge structures. They posit that all disciplined knowledge has a structure, which is reflected in the discipline’s epistemology. Without such structures, disciplines would be no more than a mass of unrelated propositions. In this understanding, Pines (1982) defines a discipline as a structured field of knowledge that encompasses a theoretical structure (i.e., conceptual framework) and a methodological structure (i.e., empirical approaches). Both conceptual and methodological structures are inseparably intertwined and should be explicitly acknowledged in the school curricula. In other words, a curricular epistemology should be the reflection of the scientific episte­mology of a discipline.

Golin (1997) also stresses the importance of the epistemic structure found in scientific knowledge to guide curriculum development. He also acknowledges the importance of educational goals and didactical

Florian C. Feucht 75

principles to inform curricular epistemologies. That is, epistemological underpinnings of curricula do not only derive epistemologies identifia­ble in scientific paradigms to be taught, but are also based on epistemic underpinnings of educational and psychological knowledge in the field of learning and instruction.

While these two authors believe curricular epistemology should more­or­less emerge from the structures of scientific knowledge, Dob­son and Dobson (1987) promote a more constructivist understanding of knowledge overall. They conceptualize both scientific knowledge and curricular knowledge as a construction of the human mind. They focus on the methodological aspects of translating scientific knowledge into school curricula considering epistemic dimensions such as sim­plicity, linearity, determination, and reversibility of knowledge. These dimensions will allow determining school subject/discipline differences on a more surface level.

Overall, there is the shared assumption that all learners should acquire an understanding of the epistemological underpinnings of the knowl­edge acquired in school subjects. Therefore, curricula should represent the epistemology of such knowledge as accurately as possible.

Curricular epistemology as worldviews. In contrast to the previous section, authors such as Benson and Griffith (1991), Hill (1973), Kilbourn (1980), and Yang (2001) propose a variety of competing curricular epistemologies that reflect different philosophical worldviews and not scientific disciplines. Interestingly, all of these authors are critical of the values and norms transported in these worldviews, as the student might be indoctrinated through the curricular epistemology. For example, Yang (2001) describes the field of curriculum and instruction as being influenced by epistemology, political and moral philosophy, and con­tent area, in a reciprocal relationship between curriculum and society. The different worldviews identifiable in curricula, for example idealism (e.g., the capacity of human beings to exercise reason) and develop­mentalism (e.g., children learn differently than do adults), are subject to individual interpretations and may result in diverse and possibly con­flicting worldviews in elementary school students. Therefore, educators need to be aware of the influence of different curricular epistemolo­gies and be able to evaluate the purpose, content, and organization of curricula.

Hill (1973) is more critical about the process of curriculum develop­ment itself. He posits that curriculum planners have the tendency to neglect the value of epistemological underpinnings in curricula in general and that their professional neutrality is severely compromised by their own personal epistemologies. Hill discusses five different

Epistemic climate in elementary classrooms76

epistemologies, including essentialism and positivism and their corre­sponding values and norms, which are reflected in curriculum design and content. For example, the positivist position is driven by the assump­tion that knowledge is supported by empirical evidence while opinions or beliefs do not meet the testability criterion. Therefore, the positiv­ist curriculum is predominantly focused on the natural and social sci­ences, which follow logic and mathematics, while other disciplines play only a subordinate, supporting role. Acknowledging the impossibility of epistemological neutrality, Hill recommends curriculum planners to follow a meta­epistemological approach (i.e., nearly value­free and non­normative position) to inform curricula theory.

Similarly, Benson and Griffith (1991) address the impact of curricula on students’ personal epistemologies. They criticize contemporary school curricula for promoting the belief in learners that knowledge is static and absolute. They describe schooling as a “process of memorizing facts or static definitions” rather than one of seeking to understand the nature of knowing (p. 25). Therefore, Benson and Griffith call for the modifica­tion of school curricula to counterbalance the static and absolutist knowl­edge representations with educational representations that are reflective of human processes of knowing. Their hope is to override the effect of epistemological mono­cultures on students’ personal epistemologies.

Kilbourn (1980), unlike the other works in this section on curricular worldviews, proposes a very different solution to overcome the influ­ence of curricular epistemology. She proposes that it is the teachers’ ethical duty to train students to develop step­by­step a sense of intellec­tual independence, allowing them to evaluate the epistemic messages of curricula on their own. Clearly, Kilburn seems to trust in the emerging epistemic abilities of elementary students. Unfortunately, no research could be identified that describes the impact of curricular epistemolo­gies on students’ personal epistemology. However, two studies exist that research such an epistemic effect on teachers.

Empirical research on curricular epistemology. Only two empirical stud­ies could be identified that address epistemological underpinnings of curricula. Both studies, conducted by Haes (1982) and Benson (1989), investigate how curricular knowledge is perceived epistemologically by teachers. Haes conducted a case study to explore to what extent teachers perceive curricular knowledge as true and certain. Among other ques­tions, teachers responded to an epistemological question about whether or not it is important for curriculum content to be true. The data analysis revealed that most of the teachers differentiated between an absolutistic, relativistic, and socially constructed understanding of truth. They per­ceived some aspects of curriculum content as scientifically absolute and

Florian C. Feucht 77

others as relativistic and/or as socially constructed. Despite this epis­temological distinction, more than half of the teachers concluded that overall curricular knowledge is true and can be known with certainty.

Benson (1989) further expanded on this matter by investigating how teachers justify their more absolutist perception of curricula. He inter­viewed and observed three biology teachers implementing a sequence of lessons on nutrition. The data analysis revealed that teachers justi­fied their view of curricula as true and certain on the bases of situ­ational constraints, such as governmental, institutional, and societal control, as well as moral and religious teaching (e.g., teachers believed that curricula are governmental documents and that they are legally bound by them). Benson concluded that this unquestioned acceptance of curricula might be indeed caused by situational constraints, but also speculated on the teachers’ lack of understanding regarding the ques­tioning of the formation of knowledge.

Summary. The theoretical literature reviewed proposed curricular epistemology as epistemological knowledge structures and as differ­ent worldviews. The first group of authors proposed that curricular epistemology should be more­or­less derived from the scientific dis­ciplines, while integration of educational principles and goals might be optional and necessary, respectively. The second group was critical about the different influences of epistemologies identifiable in curricula on students’ personal epistemologies. They suggested raising awareness of this influence in curriculum planners, teachers, and students. The two empirical studies reviewed revealed that teachers are indeed aware of the epistemic underpinnings of curricula, but refer to situational constraints to explain curricula as a certain and true representation of knowledge. Literature addressing other knowledge representations typical to elementary education, such as school books and other educa­tional materials, could not be identified.

Reciprocal relations and emerging issues

In this section all four factors of epistemic climate are integrated fol­lowing the EMPE. This integration focuses on how individual factors influence other factors including the reciprocal and dynamic relations that are essential to the nature of epistemic climate. On one side, this integrative process brings the dynamic nature of epistemic climate to life and, on the other, it further validates the conceptualization of the EMPE on the grounds of additional empirical and theoretical litera­ture. Along this process, emerging issues are discussed and a brief sum­mary will conclude the overall literate review.

Epistemic climate in elementary classrooms78

Teachers’ personal epistemology. The theoretical and empirical lit­erature on teachers’ personal epistemology demonstrates its potential impact on the other three factors of epistemic climate. For example, teachers’ selection and use of instruction is influenced by their per­sonal epistemology. Schraw and Olafson (2002), among others, provide evidence that teachers use different instruction and assessment meth­ods depending on their personal epistemology. Johnston and colleagues (2001) illustrate how the discourse pattern of two teachers, monological and dialogical discourse, differed according to their epistemic beliefs. White (2000) was able to identify five epistemologically different cat­egories of teachers based on the analysis of their instructional approach to solving classroom dilemmas.

The use of knowledge representations (e.g., curricula, school books, and educational media) also influences teachers’ personal epistemol­ogy. Schraw and Olafson (2002) theorized that curricula are differently perceived and implemented by teachers depending on their worldviews. Teachers with a realist worldview, for example, may acknowledge the existing curricula, while teachers with a relativist worldview may focus on the active and independent knowledge construction of their students. Haes (1982) and Benson (1989), who investigated different perceptions of teachers on curricula, argue that these perceptions impact their use of curricula. For example, teachers who recognize curricula as a gov­ernmental prescription presumably follow the stated learning outcomes more strictly than others (Benson, 1989). Johnston and colleagues (2001) found that teachers with different personal epistemologies value knowledge representations differently in their classrooms. The teacher who followed a monological discourse pattern made permanent use of textbooks and worksheets as they were authoritative knowledge repre­sentations, while the teacher with a dialogical discourse pattern con­sidered students’ prior knowledge and their independent knowledge construction as a more important knowledge source.

Furthermore, Johnston and colleagues (2001) demonstrated that the personal epistemology of teachers impact those of their students. Students who were exposed to the dialogical patterns of classroom discourse perceived knowing as a meaning­making activity in which their own experiences and those of their peers were important. In con­trast, students in the classrooms in which the monological approach was practiced demonstrated clear concepts of performance success and perceived themselves as passive consumers of knowledge. Johnston and colleagues’ contrasting approach illustrated that students’ personal epistemology can be linked to the personal epistemology of their teach­ers. This influence is also briefly addressed in Bendixen and Rule’s

Florian C. Feucht 79

(2004) integrative personal epistemology model. They consider teachers’ personal epistemology as an external factor, which may contribute to epistemic change in learners.

The theoretical and empirical literature on teachers’ personal episte­mology provides insights regarding its influence on the other factors of epistemic climates represented in the EMPE. More research is needed to further validate what is known about this component, but also to explore more aspects of its influence. For example, it would be of inter­est to investigate if teachers make a different use of instruction and knowledge representations across school subjects and across grade lev­els. This would shed light on whether teachers’ practice is more influ­enced by domain­general or domain­specific beliefs about knowledge. This could lead to the question, are teachers’ beliefs about knowledge and knowing implicit or explicit? Or, is their practice simply driven by accommodating basic student characteristics, such as cognitive ability and subject interest?

Learners’ personal epistemology. Little is known about the influence of learners’ personal epistemology on epistemic climate. Bendixen and Rule (2004) theorized that learners’ personal epistemology might influ­ence those of their peers and classroom teachers. Other influences have not been explicitly theorized. Interestingly, no research has yet been conducted to shed light on this issue. This lack of research might be a result of the assumption that students play a passive role in education and/or the fact that classroom­level research is not often undertaken in the field of personal epistemology (for examples, see Hammer and Elby, 2002; Feucht, 2008). Still, there is the necessity to explore the learn­ers’ impact on the epistemic climate in their classrooms. For example, looking at the diversity of students’ personal epistemology within a classroom, what are its possible side­effects on the implementation of instruction and curriculum (e.g., student interest in and perception of different instruction and educational materials)? To what extent does their personal epistemology impact those of their peers and teachers?

Epistemic instruction. More research is conducted on the influence of epistemological underpinnings of instruction on the epistemic climate. Scheffler (1965) and Wade (1975) theoretically discuss the impact of different teaching models on students’ personal epistemology. Schef­fler (1965) is quintessentially in favor of the rule model, which follows an evaluativist understanding of knowledge and knowing. Wade (1975) calls for the utilization of models, the epistemological underpinnings of which are in line with those of the subject knowledge taught. Hofer (2001) briefly describes how instructional practices and pedagogical approaches might impact learners’ personal epistemology.

Epistemic climate in elementary classrooms80

The theoretical assumptions that epistemic underpinnings of instruction can influence personal epistemologies have been somewhat verified in different studies. The explorative approaches of Louca and colleagues (2004) and Steinbring (1991) demonstrate how different instructional approaches can support the epistemic understanding of elementary school students. Boscolo and Mason’s (2001) and Smith and colleagues’ (2000) intervention studies shed light on the impact of different instruction in controlled research experiments. The results reveal that the personal epistemologies of students in the experimental groups (i.e., exposure to constructivist interventions) advance, while those in the control groups were not influenced. Results like these were found in similar intervention studies conducted with preservice teach­ers (e.g., Brownlee et al., 2001). However, these studies were conducted in teacher education and are, therefore, out of the direct influence on the epistemic climate in elementary classrooms.

A possible impact of epistemic instruction on curricula and teach­ers’ personal epistemology has not yet been the subject of theory and research. For example, could the internal consistency of epistemic dimensions in experienced teachers (Tsai, 2002) be caused by a perma­nent implementation of the very same epistemic instruction over many years? How would the epistemic underpinnings of instruction impact epistemic climate when conflicting with those of curricula and other knowledge representations?

Epistemic knowledge representations. A large body of theoretical lit­erature is published on the influence of epistemic underpinnings of curricula (i.e., curricular epistemology). Theorists defined curricu­lar epistemology either as a knowledge structure (Dobson and Dob­son, 1987; Golin, 1997; Pines, 1982) or as a philosophical worldview ( Benson and Griffith, 1991; Hill, 1973; Kilbourn, 1980; Yang, 2001). They assume that the epistemological underpinnings of curricula, as either knowledge structures or a philosophical worldview, impacts the personal epistemology of elementary students. Some authors are with­out a doubt convinced of a positive influence, while others are more critical. The latter took into account that the hidden values and norms of certain worldviews could bias students in a negative way.

Despite the existence of this large amount of theoretical literature, no research studies have been conducted to verify if and how curric­ular epistemology does actually impact the personal epistemology of students. However, two studies were identified which addressed how curricular epistemology is perceived by teachers (Benson, 1989; Haes, 1982): most of the teachers interviewed believed that curricula have a prescriptive notion which should be followed. Hence, it could be argued

Florian C. Feucht 81

that curricula might have an authoritative impact on the personal epis­temology of teachers.

To verify either of these assumptions, the impact of curricular epis­temology on the personal epistemology of students and teachers neces­sitates more research. Similarly, the extent to which epistemological aspects of curricula influences the use of instruction requires more detailed empirical research. For example, comparison studies could be conducted to explore how teachers rationalize their instructional approaches based on different curricula (e.g., idealist versus positiv­ist curricula). Most importantly, there is the need to investigate how knowledge representations, other than curricula, might contribute to the nature of epistemic climates in elementary classrooms.

Summary. Although only very little research and theory exists on the construct of epistemic climate, more literature can be identified that addresses different factors and processes that are essential in describing the nature of epistemic climate conceptualized in the EMPE. Reviewing, in particular, aspects of epistemic climate in elementary classrooms, all four factors and most reciprocal relations could be identified in accord­ance with the working definition. The reviewed theory and research supports the conceptualization of the construct and its representation in the form of the EMPE. Teachers’ personal epistemology influences learners’ personal epistemology, and their instructional choices and use of knowledge representations. In the same classroom, the epistemic underpinnings or messages of instruction and the knowledge represen­tations themselves influence students’ and teachers’ personal epistemol­ogy. Reading between the lines, it appears that most of the literature reviewed suggests a more­or­less one­directional influence of the epis­temic climate on learners’ personal epistemology. However, the working definition used in this chapter stresses elementary school children and their personal epistemology as an incremental and influential part of the epistemic climate in their classroom. While more research is needed in general to explore, measure, and further theorize epistemic climate, the equal positioning of students’ personal epistemology among the other three factors requires, in particular, the attention of future research and recommendations for classroom education.

Implications for future research and educational practice

Several issues pertinent to future research and educational implications stem from the previous literature review and discussion. Specifically, the following section focuses on educational and methodological implications

Epistemic climate in elementary classrooms82

revolving around the context­ and school subject­specificity of epistemic climates in elementary classrooms and its potential for epistemic devel­opment. The application of different conceptual frameworks, such as the EMPE and existing dimensional and developmental conceptualization of personal epistemology, are an important aspect of these implications.

The application of the educational model of personal epistemology

Overall, the EMPE can be strategically used to guide educational practices and to systematize research regarding epistemic climates in elementary classrooms. Because of its broad foundation in educational psychology and curriculum and instruction, and its strong focus on epistemic aspects of classroom education, the EMPE can be an impor­tant resource for practitioners, theorists, and researchers in many dif­ferent educational fields.

Educational implications. The EMPE can be applied in the field of teacher education and curriculum and instruction to inform teachers and other educational professionals about epistemic climates in elemen­tary classrooms (e.g., teachers’, curriculum developers’, and instructional designers’ personal epistemology). Raising their awareness of the nature of epistemic climate and the reciprocal influence among its components can positively influence the development of more appropriate curricula, school books, and other educational materials (i.e., epistemic knowledge representations) and application of different instructional approaches (i.e., epistemic instruction). For example, teachers who become aware of their own personal epistemology and learn to assess the epistemic notions underlying their students’ beliefs systems can make informed choices with regard to their instructional approaches and use of educational materials to guide students toward a more advanced and school/discipline­specific epistemological understanding. That is, a history teacher, for instance, might purposefully use educational materials (e.g., historical resources and artifacts) illustrating different perspectives on historical events to help her students gain a more multiplistic or evaluativistic understand­ing of historical knowledge. In other words, the more general conceptu­alization of the EMPE offers many opportunities to develop educational implications suitable and adoptable in a variety of different contexts and school subject areas. Still, more general research is necessary to empiri­cally inform teacher education and classroom education.

Methodological implications. Most of what is known about epistemic climate is limited to empirical investigations on one or two of its com­ponents and/or their reciprocal relations. More research is needed that

Florian C. Feucht 83

specifically explores and measures the nature of epistemic climates more comprehensively. In this line of thinking, the EMPE can be utilized to design research studies accounting for all components and relations of epistemic climate, and to systematically compare and integrate results from more focused studies into the existing body of knowledge. More specifically, the theoretical foundation of the EMPE – grounded in the fields of educational psychology and curriculum and instruction – allows taking many different and broader approaches to researching the epistemic climate from different perspectives and for different purposes. For example, the development of curricula and learning objectives can be developed in light of epistemology research, bringing together issues concerning curriculum design (e.g., content knowledge, learning objec­tives) and learners’ cognitive development (e.g., students’ level of epis­temic development). That is, intervention studies could be conducted to evaluate, for instance, what kind of curricular epistemology is most conducive to students’ epistemic development and learning at different stages of their cognitive development.

From a researcher’s perspective the EMPE can be tipped and turned around to better investigate and focus on a particular aspect of the epis­temic climate. That is, not all corners (i.e., components) and edges (i.e., reciprocal relations) of the model have yet been empirically verified. For example, little is known about how students (i.e., learners’ personal epis­temology) influence the epistemic climate surrounding them (e.g., explic­itly verses implicitly, actively verses passively). Are teachers intimidated by students who insist on their own personal views about certain topics (i.e., multiplists)? Do students avoid epistemic conflicts with their teach­ers in the pursuit of good grades? Do students’ personal epistemologies influence their teachers’ choice regarding instruction and educational materials? Furthermore, the EMPE provides sufficient conceptual sur­face to integrate and anchor points to adjoin new factors and processes, which might be of interest to researchers in different fields. For example, how are theories about self­determination, future­time­perspective, and self­regulated learning linked to the construct of epistemic climate (i.e., learners’ and teachers’ personal epistemology)? How do they influence the dynamic of epistemic climates? As a framework, the EMPE can guide future research to contribute more strategically and systematically to what little is known about epistemic climate as a classroom phenomenon.

Context- and school subject-specificity of epistemic climate

Context­ and domain­/discipline­specific aspects have been impor­tant facets of theorizing and researching personal epistemology (e.g.,

Epistemic climate in elementary classrooms84

Buehl et al., 2002; Hammer and Elby, 2002; Hofer, 2004, 2006; Muis et al., 2006). Appropriately, it can be assumed that epistemic climates are context­specific and subject/discipline­specific. Context­specificity refers to situational factors that shape epistemic climates (e.g., cultural and socio­economic background of students), while subject/discipline­specificity refers to the actual content knowledge and the epistemology of the subject and discipline, respectively (e.g., conducting scientific experiments in chemistry; comparing historical resources in social studies). In other words, epistemic climates can differ with regard to their context as well as to their subject/discipline­specificity. On the one hand, due to the contextual composition of each classroom, an epi­stemic climate can differ from classroom to classroom within the same school (e.g., different teachers, students, and instruction), between rural and urban schools (e.g., differences in students’ prior knowledge), between states (e.g., different curricula), and between different coun­tries (e.g., different educational goals and cultures) (e.g., Khine, 2008; Schommer­Aikins, 2004).

On the other hand, epistemic climates can differ according to school subject­specific content knowledge to be learned (e.g., Hammer and Elby, 2002; Louca et al., 2004; Steinbring, 1991). For example, within the same fourth­grade classroom (i.e., same teacher and students) the epistemic climate in a social studies class might be different to that in a science class. That is, the teachers’ and students’ epistemic beliefs might differ according to the subject. Furthermore, instruction and knowledge representations might be characteristic for certain subjects and differ in their epistemic underpinnings (e.g., reading different his­torical resources verses conducting scientific experiments). The class­room epistemology in a social studies class might represent knowledge more as uncertain and subjective and in a science class more as certain and objective; somewhat in accordance to their related academic disci­plines (e.g., Kattman et al., 1996; Golin, 1997; Pines, 1982). Occurring simultaneously, the context­specificity and subject­specificity of epis­temic climates clearly increase its complexity and diversity.

Educational implications. Looking at the discussion on curricular epis­temology as an indication of educational objectives, there is the position that students should form an understanding of the epistemic structure of a school subject’s corresponding discipline (i.e., epistemic climate in a science classroom should equal the epistemic climate in science), while a second position warns that students should not be biased by the philosophical worldviews (unconsciously) implanted in curricula by their developers. Essentially, a sensible combination of these posi­tions is to teach students to become critical and independent thinkers

Florian C. Feucht 85

within the epistemology of a certain discipline while acknowledging the situational context of the lesson and classroom. For example, stu­dents can learn that knowledge in literature is objective and subjective. On the objective side, there are defined criteria and certain strategies that can be used to analyze a poem, while on the subjective side, it is the readers’ judgment to decide what the analysis means to them per­sonally. Integrating both sides will allow students individually or as a class community to come to a conclusion or consensus, respectively, and therein to gain an evaluativistic understanding of the nature of literature knowledge. That is, students are explicitly introduced to the epistemological underpinnings of the literature as a subject/discipline (e.g., its established and characteristic criteria, strategies, methods, and values) and become actively involved in a critical and independent process of knowledge creation and evaluations.

To achieve this goal, teachers would need to understand the epistemic influence of their personal epistemology and their use of instruction and knowledge resources. Subsequently, they need to know how, when, and which instructional approach and knowledge representation can foster more appropriate epistemic beliefs in learners. Teachers would need to (1) have a sufficient understanding of the epistemology of a school subject and its related academic discipline, respectively, (2) be able to bring these epistemologies in sync, and (3) justify why they might choose not to bring them in sync (e.g., a scientific concept is presented in a less complex man­ner; the ambiguity of a historical event is not addressed). Finally, from a socio­constructivist perspective, teachers would also need to consider students’ personal epistemology of the school subject and the academic discipline in their lesson planning. Addressing these educational neces­sities could be subject to various forms of teacher training and develop­ment, while more research is needed to ground these empirically.

Methodological implications. Due to its subject­specific and context­specific nature, researching the epistemic climate is, without a doubt, a challenging task. There is the need for category systems to methodo­logically account for a systematic comparison and integration (i.e., data triangulation) of epistemic climates, their components, and relations within themselves and across different contexts and school subject areas. For example, the epistemic dimensions of Hofer and Pintrich’s (1997) framework can be applied as a category system (i.e., the cer­tainty, simplicity, justification, and source of knowledge) to provide additional methodological continuity to the application of the EMPE (Feucht, 2008; Haerle and Bendixen, 2008).

From a research design perspective, exploring and investigating the diversity and complexity of epistemic climates with regard to their

Epistemic climate in elementary classrooms86

context­ and school subject­specificity using case study research designs (e.g., Merriam, 1988) seems to be a promising approach. For example, the in­depth analysis of an epistemic climate within the context of a spe­cific classroom and school subject area (i.e., single case study design) could shed light not only on the influence of context and school subject on its individual components, but also how these components influence each other. Comparing epistemic climates across different classrooms, grade levels, and school subjects (i.e., multiple case study design) could over time be insightful to identify similarities and differences in their nature. Taking a step back to explore the epistemic climate in the light of cultural behavior (e.g., teachers, students, principals, and parents), artifacts (e.g., instruction, school books, curricula, and the Internet), and educational trends (e.g., focus on factual knowledge, educational accountability, and intelligent design discussions) also follows the idea of ethnographical research (e.g., Flick, 2002; Salkind, 2007) and can be embedded within socio­constructivist and systemic theories (e.g., Bronfenbrenner, 1979; Vygotsky, 1978).

Developmental aspects of epistemic climate

The literature review showed that elementary school students and teachers and personal epistemology can vary in its epistemic develop­ment, which can be described in the levels of absolutism, multiplism, and evaluativism (e.g., Kuhn et al., 2000). Similarly, the epistemologi­cal underpinnings of instruction, curricula, school books, and other educational materials can be rated along different qualitative/devel­opmental levels of epistemology. For example, teacher­centered instruction might be more absolutistic in nature, while student­centered instruction more multiplistic or evaluativistic (e.g., Schraw and Olafson, 2002; White, 2000). In other words, the developmen­tal/qualitative differences of epistemic beliefs and underpinnings of the EMPE’s four components can contribute to the overall epistemic development/quality of an epistemic climate which, in turn, might be dependent upon the epistemic characteristics of its context and school subject. For example, the epistemic climate of a math lesson might appear more absolutist reflecting the teacher’s belief that knowledge and skills in mathematics are more objective and certain, while a his­tory classroom might be more multiplistic in its nature, addressing the perceived subjectivity and uncertainty of this particular discipline. Furthermore, curricula, textbooks, and other educational materi­als might also play an important role in reflecting domain­specific

Florian C. Feucht 87

epistemologies in mathematics and history, and suggest different levels of epistemic development/quality could, in turn, also contribute to the nature of the epistemic climate.

Educational implications. Research with adolescents and adults has shown that more sophisticated epistemic beliefs (i.e., evaluativism) are more conducive to learning in general (e.g., Schommer, 1990; Kardash and Scholes, 1996). Furthermore, the previously stated goal that elementary school students should become critical and independ­ent thinkers within the epistemology of a certain discipline ties in with evaluativist beliefs about knowledge and knowing; that is, knowledge is described as an integration of objective and subjective knowledge to form tentative knowledge claims justifiable in their close (micro level) environment (Kuhn et al., 2000).1 An additional, a more general argu­ment that speaks to developing evaluativistic and critical thinking in elementary school students is that such thinking skills are essential for fostering an understanding of democracy and good citizenship even early in life (e.g., making juror decisions, forming political opinions and arguments, evaluating the agenda of politicians and their parties) (Haerle and Bendixen, 2008). In conclusion, it can be proposed that teachers and educators should cultivate evaluativistic beliefs in elemen­tary school students to help them become better learners and critical thinkers in their own right.

To achieve these goals, similar steps as described in the previous sec­tion on the context­ and school subject­specificity of epistemic class­rooms should be taken. However, in this case developmental frameworks (e.g., Kuhn et al., 2000; King and Kitchener, 1994; Perry, 1970) are center points to inform teacher education, curriculum, and textbook development. The more evaluativistic and critical thinking is fostered in epistemic climates, the more students will be equipped to accelerate in their learning and to appreciate democracy and good citizenship. In particular, a key element of teacher education should be to draw atten­tion to teachers’ epistemic development (Haerle and Bendixen, 2008)

1 Certainly, the case can be made that not all disciplinary knowledge, skills, and under­lying epistemologies are evaluativistic in their nature (e.g., some facts will be always absolute even in more subjective disciplines) and, therefore, school subjects and their content knowledge should be taught accordingly. I agree. However, I also strongly believe that any (disciplinary) scientific community that is alive and well should be practicing and teaching an evaluativistic understanding of knowledge and knowing. Hence, I propose and assume that the epistemologies of any discipline should pursue an evaluativistic level of epistemology (development). This should be reflective in the way we teach elementary students about our knowledge and skills of the culture and world surrounding us.

Epistemic climate in elementary classrooms88

and its possible supporting and/or suppressing influence on learners’ epistemic beliefs. Again, more research is needed to empirically inform the development of more detailed educational implications.

Methodological implications. Similar to the use of dimensional frame­works, developmental frameworks can be utilized to bring methodologi­cal constancy and continuity when measuring the developmental level/quality of epistemic climates, their components, and relations. Kuhn and colleagues’ (2000) developmental framework, which accounts for epistemic development in childhood and adolescence, can be opera­tionalized to rate an epistemic climate and its four components along a scale of epistemic development/qualities (Feucht, 2008). For example, the epistemic underpinnings of curricula, school books, and instruction could be rated as absolutist, multiplist, and/or evaluatist in their nature. Such ratings would not only be of empirical interest to compare different knowledge representations and their impact on the epistemic climate, its components, and relations, but also provide teachers with a tool to assess the aspects of the epistemic climate in their classrooms. Realistically, genuinely absolutist, multiplist, or evaluativist epistemic climates through­out all their components appear to be impossible due to the diversity and complexity. Nevertheless, empirical research is critical to measure the educational potential of epistemic climates in fostering and/or suppress­ing epistemic development in elementary school students.

To measure the effect of epistemic climate, their components, and influence on the development of learner’s personal epistemology, differ­ent developmental research designs are prudent, such as cross­sectional, longitudinal, and quasi­experimental studies (e.g., Salkind, 2007). Most important is micro­genetic research, which would permit infor­mation to understand change processes of epistemic beliefs and not only products of the change. That is, micro­genetic designs could be applied to investigate the micro­genesis of a moment­by­moment epis­temic change (e.g., Chinn, 2006; Lavelli et al., 2004). For example, such micro­genetic research embedded in intervention studies could shed light not only on the gradual and complex development of personal epistemology but also what kind of instruction and educational materi­als are more supportive of epistemic change in students (i.e., classroom education) and their teachers (i.e., teacher training and development). To permit research focused on epistemic development, new and/or more suitable measures are required to successfully investigate not only the complexity and dynamics of a classroom’s epistemology, its compo­nents, and relations, but also to detect and measure epistemic change among them (e.g., a portfolio of assessment tasks, checklists, structured and semi­structured questionnaires, and self­report measures).

Florian C. Feucht 89

Conclusion

Epistemic climate is a relatively new construct in the field of personal epistemology and education in general. To this point, most of what is known is grounded in literature focused on aspects of epistemic cli­mate and not on the construct as a whole. It is assumed that epistemic climate is context­ and school subject­specific and reflect different levels of epistemic sophistication. Due to its potential for influencing epistemic development, a variety of different educational implications for classroom practice can be proposed to foster evaluativist beliefs in elementary school students. These beliefs are conducive to learn­ing, reflective of disciplinary epistemologies, and promote critical and independent thinking. Researching the diversity and complexity of epistemic climates comprehensively or focusing on selected compo­nents and relations requires the operationalization of different epis­temic frameworks, such as the EMPE. This approach is essential to methodologically systematize and compare epistemic climates, their components, and relations within themselves and across different contexts, school subjects, and levels of epistemic development/quali­ties. Overall, better instruments are imperative to permit researchers and teachers to uncover the nature of epistemic climate in elementary classrooms and their potential for raising elementary school students to be evaluativist thinkers in the context of their school subjects’ epis­temology and in their future life as responsible and productive citizens in our societies.

R EF ER ENCES

Bendixen, L. D. (2002). A process model of epistemic belief change. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (191–208). Mahwah, NJ: Lawrence Erlbaum Associates.

Bendixen, L. D. and Rule, D. C. (2004). An intergrative approach to personal epistemology: A guiding model. Educational Psychologist, 39(1), 69–80.

Benson, G. D. (1989). Epistemology and science curriculum. Journal of Cur-riculum Studies, 21, 329–44.

Benson, G. D. and Griffith, B. E. (1991). The process of knowing in curricu­lum. Journal of General Education, 40, 24–33.

Boscolo, P. and Mason, L. (2001). Writing to learn, writing to transfer. In P. Tynjala, L. Mason, and K. Lonka (Eds.), Writing as a learning tool: Inte-grating theory and practice (83–104). Dordrecht, The Netherlands: Kluwer Academic Press.

Boyes, M. C. and Chandler, M. (1992). Cognitive development, epistemic doubt, and identity formation in adolescence. Journal of Youth and Adoles-cence, 21, 277–303.

Epistemic climate in elementary classrooms90

Bronfenbrenner, U. (1979). The ecology of human development. Cambridge, MA: Harvard University Press.

Brownlee, J. (2001). Epistemological beliefs in pre­service teacher education students. Higher Eduction Research and Development, 20(3), 281–91.

Brownlee, J., Purdie, N., and Boulton­Lewis, G. (2001). Changing epistemo­logical beliefs in pre­service teacher education students. Teaching in Higher Education, 6(2), 247–68.

Buehl, M. M., Alexander, P. A., and Murphy, P. K. (2002). Beliefs about schooled knowledge: Domain specific or domain general? Contemporary Educational Psychology, 27, 415–49.

Burr, J. E. and Hofer, B. K. (2002). Personal epistemology and theory of mind: Deciphering young children’s beliefs about knowledge and know­ing. New Ideas in Psychology, 20, 199–224.

Chan, K.­W. and Elliott, R. G. (2000). Exploratory study of epistemological beliefs of Hong Kong teacher education students: Resolving conceptual and empirical issues. Asia-Pacific Journal of Teacher Education, 28(3), 225–34.

(2004). Epistemological beliefs across cultures: Critique and analysis of beliefs structure studies. Educational Psychology, (24)2, 123–42.

Chandler, M. J., Hallett, D., and Sokol, B. W. (2002). Competing claims about competing knowledge claims. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (145–68). Mahwah, NJ: Lawrence Erlbaum Associates.

Chinn, C. A. (2006). The microgenetic method: Current work and extensions to classroom research. In J. L. Green, G. Camilli, and P. Elmore (Eds.), Complementary methods for research in education (439–56). Washington, DC: American Educational Research Association.

Dobson, R. L. and Dobson, J. E. (1987). Curriculum theorizing. Educational Forum, 51, 275–84.

Duit, R., Gropengießer, H., and Kattman, U. (2005). Towards science edu­cation research that is relevant for improving practice: The model of educational reconstruction. In H. E. Fischer (Ed.), Developing standards in research on science education. The ESERA Summer School 2004 (1–9). London: Taylor and Francis.

Elder, A. D. (2002). Characterizing fifth grade students’ epistemological beliefs in science. In B.K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing (347–63). Mahwah, NJ: Lawrence Erlbaum Associates.

Feucht, F. C. (2008). The nature of epistemic climates in elementary education. Unpublished doctoral dissertation, The University of Nevada, Las Vegas.

Feucht, F. C. and Bendixen, L. D. (in press). Exploring similarities and differ­ences in personal epistemologies of US and German elementary school teachers. Cognition and Instruction.

Flick, U. (2002). An introduction to qualitative research. Thousand Oaks, CA: Sage.

Gill, M. G., Ashton, P. T., and Algina, J. (2004). Changing preservice teachers’ epistemological beliefs about teaching and learning in mathematics: An intervention study. Contemporary Educational Psychology, 29(2), 164–85.

Florian C. Feucht 91

Golin, G. (1997). Structure of scientific knowledge and curriculum design. Interchange, 28 (2–3), 159–69.

Haerle, F. C. (2006). Personal epistemologies of fourth graders: Their beliefs about knowledge and knowing. Oldenburg, Germany: Didaktisches Zentrum.

Haerle, F. C. and Bendixen, L. D. (2008). Personal epistemology in elementary classrooms: A conceptual comparison of Germany and the United States and a guide for future cross­cultural research. In M. S. Khine (Ed.), Knowing, knowledge and beliefs: Epistemological studies across diverse cultures (165–90). Dordrecht, Netherlands: Springer Verlag.

Haes, J. (1982). Conceptions of the curriculum: Teachers and “truth”. British Journal of the Sociology of Education, 3, 57–76.

Hammer, D. H. and Elby, A. (2002). On the form of personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (169–90). Mahwah, NJ: Lawrence Erlbaum Associates.

Hill, B. V. (1973). Epistemologies and curriculum models. Journal of Educa-tional Thought, 7(3), 151–64.

Hofer, B. K. (2001). Personal epistemology research: Implications for learning and teaching. Educational Psychology Review, 13, 353–83.

(2004). Exploring the dimensions of personal epistemology in differing classroom contexts: Student interpretations during their first year of col­lege. Contemporary Educational Psychology, 29, 129–63.

(2006). Beliefs about knowledge and knowing: Integrating domain specifi­city and domain generality. Educational Psychology Review, 18, 67–76.

Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learn­ing. Review of Educational Research, 67, 88–140.

(Eds.) (2002). Personal epistemology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

Howard, B. C., McGee, S., Schwartz, N., and Purcell, S. (2000). The expe­rience of constructivism: Transforming teacher epistemology. Journal of Research on Computing in Education, 32(4), 455–65.

Johnston, P., Woodside­Jiron, H., and Day, J. (2001). Teaching and learning literate epistemologies. Journal of Educational Psychology, 93(1), 223–33.

Kardash, C. M. and Howell, K. L. (2000). Effects of epistemological beliefs and topic­specific beliefs on undergraduates’ cognitive and strategic process­ing of dual­positional text. Journal of Educational Psychology, 92, 524–35.

Kardash, C. M. and Scholes, R. J. (1996). Effects of preexisting beliefs, episte­mological beliefs, and need for cognition on interpretation of controversial issues. Journal of Educational Psychology, 88, 260–71.

Kattmann, U., Duit, R., Gropengießer, H., and Komorek, M. (1996). Edu­cational reconstruction – bringing together issues of scientific clarifica­tion and students’ conceptions. Paper presented at the Annual Meeting of the National Association of Research in Science Teaching (NARST), St. Louis, MI.

Khine, M. S. (2008). Knowing, knowledge and beliefs: Epistemological studies across diverse cultures. Dordrecht, Netherlands: Springer.

Kilbourn, B. (1980). World views and curriculum. Interchange on Educational Policy, 11(2), 1–10.

Epistemic climate in elementary classrooms92

King, P. M. and Kitchener, K. S. (1994). Developing reflective judgment. San Francisco, CA: Jossey­Bass.

Kitchener, R. F. (2002). Folk epistemology: An introduction. New Ideas in Psy-chology, 20, 89–105.

Kuhn, D., Cheney, R., and Weinstock, M. (2000). The development of episte­mological understanding. Cognitive Development, 15, 309–28.

Lavelli, M., Pantoja, A. P. F., Hsu, H. C., Messinger, D., and Fogel, A. (2004). Using microgenetic designs to study change processes. In D. G. Teti (Ed.), Handbook of research methods in developmental science (40–65). Cam­bridge: Blackwell Publishers.

Louca, L., Elby, A., Hammer, D., and Kagey, T. (2004). Epistemological resources: Applying a new epistemological framework to science instruc­tion. Educational Psychologist, 39(1), 57–68.

Maggioni, L., Riconscente, M. M., and Alexander, P. A. (2006). Perceptions of knowledge and beliefs among undergraduate students in Italy and in the United States. Learning and Instruction, 16(5), 467–91.

Mansfield, A. F. and Clinchy, B. (2002). Toward the integration of objectivity and subjectivity: Epistemological development from 10 to 16. New Ideas in Psychology, 20, 225–62.

Merriam, S. B. (1988). Case study research in education: A qualitative approach. San Francisco: Jossey­Bass.

Muis, K. R., Bendixen, L. D., and Haerle, F. C. (2006). Domain­generality and domain­specificity in personal epistemological research: Philosophi­cal and empirical reflections in the development of a theoretical frame­work. Educational Psychology Review, 18, 3–54.

Murphy, P. K., Alexander, P. A., Greene, J. A., and Edwards, M. N. (2007). Epistemological threads in the fabric of conceptual change research. In S. Vosniadou, A. Baltas, and X. Vamvakoussi (Eds.), Reframing the con-ceptual change approach in learning and instruction (105–22). Oxford, UK: Elsevier.

Palmer, B. and Marra, R. M. (2008). Individual domain­specific epistemolo­gies: Implications for educational practice. In M. S. Khine (Ed.), Knowing, knowledge and beliefs: Epistemological studies across diverse cultures (325–50). New York: Springer.

Patrick, H. and Pintrich, P. R. (2001). Conceptual change in teachers’ intui­tive conceptions of learning, motivation, and instructions: The role of motivational and epistemological beliefs. In B. Torf and R. Sternberg (Eds.), Understanding and teaching the intuitive mind (117–43). Mahwah, NJ: Lawrence Erlbaum Associates.

Perry, W. G. Jr. (1970). Forms of intellectual and ethical development in the college years: A scheme. New York: Holt, Rinehart and Winston.

Pines, A. L. (1982). Curriculum development and instructional planning within an epistemological­psychological framework: A theoretical synthe­sis. Journal of Curriculum Theorizing, 4(1), 88–105.

Qian, G. and Pan, J. (2002). A comparison of epistemological beliefs and learn­ing from science text between American and Chinese high school stu­dents. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The

Florian C. Feucht 93

psychology of beliefs about knowledge and knowing (389–414). Mahwah, NJ: Lawrence Erlbaum Associates.

Salkind, N. (2007). Exploring Research. New York: Prentice Hall.Scheffler, I. (1965). Philosophical models of teaching. Harvard Educational

Review, 25, 131–43.Schommer, M. (1990). Effects of beliefs about the nature of knowledge on

comprehension. Journal of Educational Psychology, 82, 498–504.Schommer­Aikins, M. (2004). Explaining the epistemological belief sys­

tem: Introducing the embedded systemic model and coordinated research approach. Educational Psychologist, 39, 19–29.

Schraw, G. and Olafson, L. (2002). Teacher’s epistemological worldviews and educational practices. Issues in Education, 8(2), 99–148.

Sinatra, G. M. and Kardash, C. A. M. (2004). Teacher candidates’ epistemo­logical beliefs, dispositions, and views on teaching as persuasion. Contem-porary Educational Psychology, 29(4), 483–98.

Smith, C. L., Maclin, D., Houghton, C., and Hennessey, M. G. (2000). Sixth­grade students’ epistemologies of science: The impact of school science experiences on epistemological development. Cognition and Instruction, 18(3), 349–422.

Steinbring, H. (1991). The concept of chance in everyday teaching: Aspects of a social epistemology of mathematical knowledge. Educational Studies in Mathematics, 22(6), 503–22.

Tsai, C.­C. (2002). Nested epistemologies: Science teachers’ beliefs of teach­ing, learning and science. International Journal of Science Education, 24(8), 771–83.

Vygotsky, L. S. (1978). Mind and society: The development of higher psychological processes. Cambridge, MA: Harvard University Press.

Wade, S. H. (1975). Epistemology and the matching of intentions with models of religious teaching. Religious Education, 70(3), 227–34.

Walton, M. D. (2000). Say it’s a lie or I’ll punch you: Naïve epistemology in classroom conflict episodes. Discourse Processes, 29(2), 113–36.

Westphal, E. (1990). Unterricht und Leben: Zur Theorie und Praxis lebensprob-lemzentrierter Unterrichtsgestaltung [Lessons and life: About theory and practice of life-problem-centered lesson planning]. Oldenburg: Zentrum für Pädago­gische Berufspraxis.

White, B. C. (2000). Pre­service teachers’ epistemology viewed through per­spectives on problematic classroom situations. Journal of Education for Teaching, 26(3), 279–306.

Yang, O. S. (2001). An epistemological and ethical categorization of per­spectives on early childhood curriculum. International Journal of Early Childhood, 33(1), 1–8.

94

4 The integrative model of personal epistemology development: theoretical underpinnings and implications for education

Deanna C. Rule and Lisa D. BendixenUniversity of Nevada, Las Vegas

Epistemological beliefs are an important aspect of current psychologi­cal thinking. Personal epistemological development and epistemological beliefs involve: how individuals come to know, the theories and beliefs they hold about knowing, and the manner in which such epistemologi­cal premises are a part of and an influence on the cognitive processes of thinking and reasoning (Hofer and Pintrich, 1997, p. 88).

In the classroom, students are exposed to new ideas, theories, and information from different sources every day. Students’ beliefs about knowledge and knowing influence the way they perceive and respond to the learning situations they encounter. For example, as students study an historical event such as the US civil war, they may have the opportu­nity to examine different sources of information such as a textbook that describes the war, personal letters written by soldiers, and/or analyses of the war published by history experts. The views of the students and other aspects of the epistemic climate (Bendixen and Rule, 2004) of their learning environment influence how these students will proceed. For instance, learners may be passive or active in this process, they may be asked to weigh evidence about these different historical sources and make their own warranted judgments, or they may be expected to find the one correct version of what happened and not question it. In addi­tion, students may think that historical information is more relative and open to interpretation while a subject such as mathematics leaves very little room for student interpretation.

Goals of the chapter

How do learners come to have certain epistemological beliefs? How could beliefs shift over time and what is the role of the teacher, for

Deanna C. Rule and Lisa D. Bendixen 95

example, in this development? Is epistemic belief change an arduous process? How can researchers better understand the construct of per­sonal epistemology and its growth? Insight into important questions such as these is at the heart of the current chapter. With this goal in mind, we will delve deeper and theoretically examine certain compo­nents of the integrative model (IM) for personal epistemological devel­opment (Bendixen and Rule, 2004) to illuminate its implications for research and education.

By examining some of the theoretical roots of the IM we hope to further clarify the model specifically and to add to our understanding of epistemological development more generally. In doing so we may reconsider the boundaries that shape the construct of personal episte­mology. As a result, the borders may change form and at times borrow from other constructs until those tentative precursors become a part of a more clarified construct. Taking this perspective in future theory and research will also enable researchers to actively influence the lim­its that outline personal epistemology and determine the profile of the construct. Finally, how the IM pertains to learning and instruction can also become more apparent as we more clearly understand the proc­esses that support epistemological development. Therefore, we seek to thoroughly consider and discuss the many significant educational implications that can be drawn from this work.

In support of our purposes, the chapter includes four sections that: (1) briefly describes the personal epistemology development literature, (2) describes the integrative model, (3) examines more deeply three spe­cific aspects of the IM (i.e., epistemic volition, reciprocal causation, and equilibration), and (4) includes educational implications and examples stemming from the IM.

Personal epistemology development

Over the past thirty years, a body of research has developed that has allowed researchers to address issues related to our beliefs about knowledge and knowing and their relationship to learning and stu­dent acheivement. This body of research has been pursued and ana­lyzed under approaches with disparate names (Hofer, 2004), including epistemological beliefs (Kardash and Scholes, 1996; Qian and Alver­man, 2000; Schommer­Aikins, 2004), reflective judgment (King and Kitchener, 1994), ways of knowing (Belenky et al., 1986), epistemologi­cal reflection (Baxter Magolda, 2001), epistemological theories (Hofer and Pintrich, 1997), epistemic beliefs (Bendixen, 2002), and epistemo­logical resources (Hammer and Elby, 2002).

The integrative model of personal epistemology96

These disparate approaches, although distinct in their conceptions of personal epistemology, pursue the same goal of addressing beliefs about knowledge and knowing and their relationship to learning. A review of personal epistemology research reveals that there are several approaches to conceptualizing beliefs about knowledge and knowing and their rela­tionship to learning. These approaches have uncovered that students’ beliefs about knowledge and knowing have both direct and indirect influences on student achievement (Hofer and Pintrich, 1997, 2002; Muis et al., 2006). One substantial body of research has been devel­opmental in nature and considers a general patterned sequence in the development of these beliefs (Baxter Magolda, 1992; Belenky et al., 1986; King and Kitchener, 1994; Kuhn, 1991; Perry, 1970). The origins of epistemic development are beginning to be investigated (e.g., Burr and Hofer, 2002) and the patterns of epistemic thinking are not neces­sarily considered to be strictly linear and only related to college­aged students (e.g., Chandler et al., 2002; Haerle, 2006). Thus, investigating the developmental patterns of epistemic thinking remain an important area for the field in terms of theory, research, and education.

For the purposes of the current chapter, we briefly describe the devel­opmental theory of Kuhn and colleagues (e.g., Kuhn and Weinstock, 2002) in particular to highlight and discuss the developmental pat­terns that are central to all of the developmental theories associated with personal epistemology. In general, the pattern of epistemological development has been described as occurring in three distinct forms of thinking about the nature of knowledge and the process of know­ing. In the first form (i.e., objective), views about knowledge are very simple and dichotomous; truth is judged based on an objective, exter­nal reality. The relativistic nature of knowledge (i.e., subjective) is the focus in the second form of thinking where each claim is considered equally legitimate and, therefore, cannot be judged beyond mere opin­ion. The third form of epistemic thinking integrates the objective and subjective nature of knowledge (i.e., evaluativism) and considers how differing viewpoints can be judged based on established criteria (Kuhn and Weinstock, 2002). It is important to note that evaluativistic views of knowledge are qualitatively distinct from the previous two forms of thought. More research is needed to determine how this integration of objectivity and subjectivity in knowledge takes place and the value of it in learning and instruction.

Researching how individual conceptions of knowledge and knowing develop through a process of equilibration is central to Piaget’s (1950, 1970, 1985) work and is the underlying theory for understanding epis­temological advancement (Bendixen, 2002; Hofer and Pintrich, 1997).

Deanna C. Rule and Lisa D. Bendixen 97

Piaget is responsible for generally influencing much of the research in personal epistemology along these lines. For example, many of the developmental theories (e.g., Baxter Magolda, 2004; King and Kitchener, 1994) refer indirectly to a mechanism of change but do not describe it in any great detail. For example, King and Kitchener (1994) consider reflective judgment development to occur when the beliefs about knowledge that a person is operating under do not coincide with experiences in the environment. Thus, in response to these contradic­tions, prior assumptions need to be examined and possibly changed.

Several questions come to mind in pursuit of understanding epis­temological development. How are beliefs about knowledge defined, acquired, and developed? What is the mechanism of change that takes place in their development? Are beliefs that are related to knowledge and knowing permanent once constructed? Can furthering our under­standing of a mechanism of epistemological change enhance learning and instruction? With these questions and others in mind regarding epistemological development, we turn to the next section of the chap­ter that will briefly describe the specific components of the integrative model.

The integrative model for personal epistemology development

In proposing the integrative model we intended to provide some guid­ance for theory and research in the field of personal epistemology (see Figure 4.1). For the purposes of the current chapter we also hope to bring the IM more explicitly into the educational arena. In this sec­tion we provide a fictitious classroom example as a vehicle to briefly describe the IM and its components (for a more detailed description of the IM’s components, see Bendixen and Rule, 2004).

Mr. Dyson’s classroom

Mr. Dyson teaches fourth grade at a public elementary school. He has just introduced a new teaching protocol in mathematics. On this partic­ular day he decides to place large pieces of blank white paper over several desks and instructs the students to solve their math problems and put their work directly on the paper so others can see. During this process he then suggests that each student think about his or her own solution and compare the process with other students. Mr. Dyson then urges the students to think about and decide which method is the preeminent method if any particular “best” method exists. At first, several of the

The integrative model of personal epistemology98

students giggle and seem uneasy. One of these students asks, “How can we come up with different ways to solve a math problem? Isn’t there only one way to get to the answer?” Other students dive right into the task and seem excited about being let loose to solve the problem on their own. As the activity progresses many of the students are fully engaged in sharing and critiquing the different approaches to the problem and there are many “Aha” moments. In the end, the class decides that there are two different yet viable ways to solve the problem. The other approaches that were tried were not as useful. A small number of stu­dents still seem uncomfortable and unconvinced about the process. One of the skeptical students mumbles, “That was dumb, why couldn’t he just tell us the right way and then we could skip all of that other stuff; he’s the teacher after all!” Mr. Dyson then asks the students to journal about their experiences that day and to think about their own process of learning. Overall, Mr. Dyson is excited about how the day’s activities went and, in particular, is pleasantly surprised at the high level of ques­tioning that came from most of the students as compared to previous days. He plans to incorporate activities like these in all of the subjects he is responsible for, such as social studies and literature.

Epistemicdoubt

Epistemicvolition

Resolutionstrategies

Advancedbeliefs

Currentbeliefs

Cognitiveabilities

Cognitiveabilities

Personal epistemologymultiplier

Affect

Environment

Mechanism of change

SoK

JKSK

CK

Metacognition

Conditions for change

Reciprocalcausation

Figure 4.1: The integrative model for personal epistemology devel­opment (IM) (Bendixen and Rule, 2004) with the addition of the personal epistemology multiplier (Rule, 2003)

Deanna C. Rule and Lisa D. Bendixen 99

The integrative model

Much of what happened in Mr. Dyson’s classroom can be considered in terms of the integrative model. In the following section we will describe various aspects of the IM (i.e., mechanism of change, reciprocal caus­ation, dimensions of beliefs, meta­cognition, affect, environment, and equilibration) and use Mr. Dyson as our example. This approach is also in line with our goal to elaborate on the educational potential of the IM.

Mechanism of change. The focal point of the IM is the mechanism of change and it consists of three interrelated components: (1) epis­temic doubt, (2) epistemic volition, and (3) resolution strategies (see Figure 4.1). Hence, there is one operating mechanism with three com­ponents as opposed to three individual mechanisms.

The first component of the mechanism of change within the IM focuses on epistemic doubt. This is viewed as a specific form of cogni­tive dissonance associated with questioning one’s beliefs about know­ledge and knowing. Epistemic doubt alone does not imply a progression through the mechanism of change; therefore, epistemological change is not always a given even if epistemic doubt is experienced (Bendixen, 2002; Bendixen and Rule, 2004). Figure 4.1 depicts an arrow moving from doubt back to current beliefs, suggesting this reversion to original beliefs is a possibility. Students in Mr. Dyson’s classroom showed clear evidence of experiencing some aspects of epistemic doubt in their ques­tions (e.g., “Why do we have to do this?”) and in some of their behaviors (e.g., frustration when a problem­solving approach was not successful). In other students, epistemic doubt may have been more subtle and brief and, therefore, not as apparent in their outward behavior.

Epistemic doubt can be part of the impetus for epistemic change, but, according to the IM, change also requires epistemic volition (see Figure 4.1). The construct of volition has been associated with the motivation and conceptual change literature. Corno (1993) defines volition as a “dynamic system of psychological control processes that protect concentration and directed effort in the face of personal and/or environmental distraction” (p. 16). Volition is critical to the IM in that it focuses on the individual taking “responsibility” for their epistemo­logical beliefs (Baxter Magolda, 2004). Epistemic volition is a powerful contributor to epistemic change; not only does change involve epistemic doubt, change requires action.

Was epistemic volition evident in the problem­solving of Mr. Dyson’s students? Mr. Dyson certainly provided the students the opportunity to “take charge” of their own beliefs and learning. Solving the prob­lem more on their own and the need to make judgments and choices

The integrative model of personal epistemology100

regarding their own beliefs (and others) was apparent. We assume, therefore, that some of the students did in fact take volitional control of their epistemic beliefs and this led to a deeper understanding of mathematics.

The third component of the mechanism of change in the IM includes the resolution strategies one may implement to obtain belief change, and which are dependent upon previously experiencing epistemic doubt and volition. As was stated, resolution strategies can facilitate a progression to advanced beliefs, but it is important to note that a reversion back to current beliefs still remains a distinct possibility (see Figure 4.1). Resolution strategies can include reflection, social interaction and, more generally, retrospective review, belief implications analysis, and educated choices (Baxter Magolda, 2004; Bendixen, 2002).

What were some of the resolution strategies the students were exhibiting in Mr. Dyson’s class? Consistent with the IM, students were interacting with their classmates (peers) to find closure to some of their questions. The majority of the class came together as a whole to make final judgments about what approaches were more useful. This speaks to potential avenues for resolution that might occur in a classroom setting. Mr. Dyson may also want to check the students’ journals to see if his students describe other aspects of their attempts at resolution.

Reciprocal Causation. The reciprocal effects of the IM’s components play a prominent role (see Figure 4.1). It is not only the individual’s per­sonal epistemology that influences his or her environment but also the personal epistemologies of others with whom that individual comes into contact. This aspect of the IM is consistent with Schommer­Aikins’ (2004) embedded systemic model of the study of epistemological beliefs that includes “feedback loops” (i.e., individuals and their environments reciprocally influencing each other). More specifically, she proposes that if a teacher encourages critical thinking and evaluation of expert assertions, students may revise their epistemological beliefs. As was illustrated in Mr. Dyson’s class, this enables students to question the teacher’s authority as absolute and become more active as learners. In turn, such a change could potentially initiate an epistemic multiplier effect (Rule, 2003) at the individual and classroom level resulting in poten­tially large changes in epistemological development (see Figure 4.1).

Dimensions of Beliefs. The IM also considers beliefs about knowl­edge and knowing to be multidimensional and in line with Hofer and Pintrich’s (1997) dimensions (i.e., simplicity of knowledge, certainty of knowledge, justification of knowledge, and source of knowledge). Some of the students in Mr. Dyson’s classroom were considering different

Deanna C. Rule and Lisa D. Bendixen 101

aspects of knowledge such as viewing mathematics knowledge as simple and certain and coming up with ways to justify their approach to solv­ing the problem as better than others. As can be seen in Figure 4.1, pro­gression through the model leads to more advanced beliefs, although there are no guarantees.

Meta-cognition. Meta­cognition is another important aspect of the IM in that the more students can explicitly consider their beliefs and learning, the more chance there is for lasting improvement. Meta­cognitive pro­cesses are also significant in Hofer’s (2004) conceptualization of personal epistemology as epistemic meta­cognition. For example, in the classroom when determining authority credibility, not only is meta­cognitive reflec­tion important, but the students’ engagement in deciding which author­ity is credible plays a role in epistemological development. Mr. Dyson was encouraging meta­cognition, for example, through the students’ use of the blank paper, open discussion of their thought processes related to the math problem, and journaling.

Affect. The significant role of emotions in epistemological develop­ment is also a facet of the IM. It is interesting to consider in what ways affect may constrain and/or facilitate epistemological development. Some students in Mr. Dyson’s class were excited about what they were doing and fully engaged in the task, while others seemed to let their more negative, unsure feelings prevent them from trying this new way of problem­solving. There have been recent calls for research in the field of personal epistemology to examine how affect influences episte­mological development (e.g., Bendixen, 2002; Pintrich, 2002).

Environment. As can be seen in Figure 4.1, the broader environment, or epistemic climate (i.e., classroom conditions that influence episte­mological development) in which the students are operating influences epistemological development. There are, of course, many aspects of environmental influences; the role of peers is one important feature of the IM. Mr. Dyson relied heavily on peer interaction and this seemed to lessen his role as absolute authority and to increase the value of con­sidering varying viewpoints rather than students relying on just his own authority as the teacher.

Equilibration. In general, the IM is a dynamic model based on an equilibration process similar to Piagetian theory (see Figure 4.1). Piaget (1985) considered the equilibration process to be the driv­ing force behind an individual’s cognitive development. Through the ongoing processes of accommodation (i.e., changing existing schemes to fit new information encountered) and assimilation (i.e., incorporating new information into existing schemes), the process of equilibration leads to revisions in cognitive systems over time.

The integrative model of personal epistemology102

This idea of development parallels the IM’s mechanism of change (i.e., an individual experiencing epistemic doubt and resolution) and also some of the IM’s broader processes (e.g., reciprocal causation). We can assume that a few of the students in Mr. Dyson’s classroom reached some form of equilibrium when their problem­solving strate­gies were successful, and what happened matched or added to their epistemological beliefs about mathematics. Other students may not have worked through their disequilibrium yet and might need further attention from Mr. Dyson.

Implications

Using Mr. Dyson’s classroom as an illustration of the IM offers some practical clarity regarding its components and certainly conjures up many possibilities concerning its implications for research, learning, and instruction. For example, it may be that epistemological develop­ment is not always about the major, sweeping “wholesale” changes in beliefs about knowledge and knowing that are often discussed in the literature (Chandler et al., 2002). The Mr. Dyson scenario showcases the daily more subtle interaction that can take place between teachers and students and students and their peers. Although more subtle, these dealings with the nature of knowledge and knowing may be quite influ­ential and important. Some of the IM’s implications along these lines will be discussed in the current chapter.

The keystones of the integrative model

In the following section we have chosen to focus on three facets of the IM and consider their theoretical roots. The keystones are epistemic volition, reciprocal causation, and equilibration (briefly described previously) (see Figure 4.1). We call these facets keystones because they are very important aspects of the IM and they offer a number of implications for theory, research, and learning and instruction. First, we examine epistemic voli­tion and its pivotal role in personal epistemological advancement. Second, reciprocal causation is considered and how it may support epistemologi­cal belief change. Third, equilibration is discussed as an important influ­ence on the IM’s mechanism of change and the model overall.

The theoretical underpinnings of epistemic volition

In general, the concept of volition draws upon current and more historical areas within educational psychology. We want to highlight

Deanna C. Rule and Lisa D. Bendixen 103

and delve deeper into the epistemic volition component of the IM because we think that it has many implications for furthering epistemic development. Indeed, we think that epistemic volition is the key to unlocking the door to personal epistemological advancement. As was discussed previously, epistemic volition is the action an individual takes in response to epistemic doubt towards resolution (see Figure 4.1). It is not enough to doubt one’s epistemic beliefs; action is required as well. We want to build the case that epistemic volition must be conscious and efficient to have any significant impact on lasting epistemic belief change. Other versions of epistemic volition may occur but we will focus on the kind of epistemic volition that we think is most conducive to lasting change. To frame our arguments for the importance of epis­temic volition we examine its links to schema theory, whether volition is conscious or automatic, and consider it in terms of the conceptual change literature in regards to intentionality.

Epistemic volition and schema theory. Reviewing aspects of Schema Theory (Anderson et al., 1977) allows us to understand the human mind playing an important role in epistemic volition and subsequent personal epistemological advancement. This may be explained by our ability to frame and organize incoming information during the knowl­edge construction process. Schemata, previously acquired knowledge stored in our memory, allow us to interact with incoming information to form new and different interpretations (Reynolds et al., 1996). It is here where we see the mind playing a more interactive role with experi­ence. The mind is playing an interactive role, but an emphasis is placed on the internal processes of information representation, organization, and framing rather than on experience. If the mind plays an interac­tive role with experience in the knowledge acquisition process, then knowledge can be represented in a fluid, dynamic way (Reynolds et al., 1996).

Anderson et al. (1977) determined that high­level schemata provide the interpretive framework for comprehending discourse and that high­level schemata influence people to see messages in certain ways. Hence, high­level schemata can influence an individual to impose a framework on a message without that individual recognizing this process. If this is a possibility in comprehending discourse, the same may be true for the mechanism of change in the IM’s consideration of personal epis­temological advancement. For example, what occurs when a person doubts their epistemological beliefs? How does an individual’s schema come into play? In terms of text­processing, for example, intrusions may first appear and ambiguous material is distorted to align the mes­sage and subsuming schemata. Second, distortions and intrusions may

The integrative model of personal epistemology104

occur only where there is no alignment between the schemata in read­ing the text and the schemata that the individual utilized to assimilate the material (Anderson et al., 1977).

When an individual doubts his or her existing epistemological beliefs, he or she may begin to question. At this point, that individual may take volitional control of his or her doubt and try to resolve it if that is a part of their background knowledge in problem­solving. The result can be in the formation of resolution strategies or perhaps the alignment of existing beliefs and schemata. What influence does that individual’s schema have on his or her own personal epistemological advancement? Perhaps schemata or background knowledge provide the interpretive framework for comprehending an individual’s own personal epistemo­logical advancement, but it may not always operate independently.

Schemata provide individuals with the interpretive framework or background knowledge necessary in the knowledge construction proc­ess, but volitional control may sometimes be required to resolve distor­tions with the alignment of existing beliefs and schemata. For example, if a student has schemata consistent with evaluativist thinking (i.e., weighing evidence and making informed choices), a distortion or dif­ference in thought will lead to that individual taking volitional control of his or her doubt because that is an essential part of evaluativist think­ing. It could be that for this student, this approach has been encour­aged by teachers and has been successful for them in past learning. Thus, this evaluativist schemata for learning that includes volition is firmly embedded and relied upon when encountering new information in the environment. Some of the students in Mr. Dyson’s classroom, for example, may experience a misalignment and distortion in terms of their current beliefs/schemata and this new way of doing math. Tak­ing volitional control of the situation may be foreign to some of the students while others are able to accomplish this task more easily and successfully.

Is epistemic volition conscious or automatic? Additional important ques­tions regarding epistemic volition involve whether it is a conscious or automatic process. For example, how does volition interact with the conscious human mind and the ability to acquire and represent knowl­edge? How do we make the distinction between automatic and con­scious processes? Automatic processes become automatic when they require little attention to be engaged, and conscious processes require increased amounts of attention because of the inherent importance of completing the task (Reynolds and Sinatra, 2009). Kahneman (2003) discusses a two­system view of cognition and illustrates the distinction between intuition (automatic) and reasoning (conscious) processing.

Deanna C. Rule and Lisa D. Bendixen 105

As Kahneman states, “there is considerable agreement on the char­acteristics that distinguish the two types of cognitive processes, which Stanovich and West (2000) labeled System 1 and System 2” (p. 698). For example, System 1 operations can be described as fast, parallel, auto­matic, effortless, implicit, and emotional. Whereas System 2 operations are different, as they require slow, serial, controlled, effortful, and rule­governed cognitions. The question remains, when does an individual select intuition over reasoning? When does an individual choose to reason? Does this involve volition? According to Kahneman (2003), a central finding in studies of intuitive decisions is that experienced decision­makers working under pressure (e.g., firefighter captains) often perceive a single option only as opposed to several options. Other options are simply not represented.

This leads us to selective attention which is critical in understand­ing volition as it is applied to the IM. Where does volition come from? Selective attention sets the stage for this answer. Is volition simply a conscious choice? Not necessarily. Conscious choice is a theme that runs through the educational psychology literature, schema theory, and volition. In order for volition to take hold as described in the mechan­ism of change proposed in the IM efficiency needs to take place. As will be discussed in a later section, Vosniadou (2003) also discusses the importance of efficiency in conceptual change learning. For example, selective attention is the ability to attend to some sensory inputs while simultaneously not attending to others, which is consistent with an experience­centered epistemological view of the mind (Reynolds and Sinatra, 2009; Reynolds et al., 1996). When processes become more automatic, we become more efficient in the way that we make decisions. The more efficient we become, the more able we are to consciously address issues that require more effortful and deliberate cognitions. These effortful and deliberate cognitions are what we can define as volition, especially as it pertains to epistemic volition as volition that is consciously manifested and efficient. This is demonstrated in the IM’s mechanism of change where conscious and efficient epistemic volition can lead to resolution and change in personal epistemological beliefs.

Similarly, in Mr. Dyson’s classroom, efforts are made to make the mathematics problem­solving taking place more process­oriented and transparent for the students. This may, in turn, allow the students to deliberate on their thought processes and decision­making. We argue that these teaching techniques would increase the chances of epistemic volition to take place.

Epistemic volition and intentionality. We draw upon the conceptual change literature to further our examination of epistemic volition. In

The integrative model of personal epistemology106

particular, we see many important parallels between its more recent consideration of intentionality and epistemic volition. For example, Mason’s body of work (e.g., 2002, 2003) reflects the interconnected nature of epistemic beliefs and intentional conceptual change learn­ing. She has found evidence to support that epistemic beliefs may act as productive resources or constraints in the conceptual change proc­ess. For example, part of lasting conceptual change is the student’s ability to recognize that there is a problem in their knowledge of a certain concept. According to Mason, more advanced epistemological beliefs are more conducive to this recognition and these students are also more likely to invest the effort in solving knowledge problems. The effortful and intentional aspect of students’ conceptual change learning that Mason describes is consistent with our view of epistemic volition as well.

In line with our previous discussions of epistemic volition, Sinatra and Pintrich (2003) define intentional conceptual change as a goal­directed and conscious process. In addition, Vosniadou (2003) argues that not all learning is intentional. Much of the learning that takes place in a typical classroom is done on an implicit level. This type of learning, she argues, learning that is non­critical and lacking in aware­ness, may be less than adequate and unstable in terms of conceptual change. Learning that is intentional (i.e., more purposeful, planful, and metacognitive), on the other hand, can greatly promote conceptual change. Learning complex concepts is “not just the simple assimilation of new information into existing structures” without being intentional. “Intentional learning can greatly facilitate conceptual change not only by making the monitoring more efficient, but also by making learners have greater metaconceptual awareness of the underlying beliefs and presuppositions and access to greater and more efficient mechanisms of the acquisition of knowledge” (p. 377).

We think there are a number of links among our view of epistemic volition and the intentional conceptual change literature. As in fruit­ful conceptual change, epistemic volition must be effortful, metacogni­tive, and purposeful for shifts in epistemic beliefs to occur. This has a number of educational implications that will be discussed in a later section.

Conclusion. As discussed previously, we have drawn upon various considerations of volition in the educational psychology literature to further understand it in relation to our IM. Consistent with the lit­erature, for epistemic volition to have the possibility of lasting belief change, it needs to be conscious and efficient. Efficiency implies the level of consciousness that is present in the volition that is exhibited.

Deanna C. Rule and Lisa D. Bendixen 107

Hence, the more conscious the volition, the more efficient and effective it will be.

When applying this to personal epistemology we can see that an increase in epistemic volition might impact personal epistemological advancement. Epistemic volition may also lead the way to resolution strategies that are consistent with the IM. For instance, volition allows the individual to take charge, make a conscious choice, and adhere to the goal of resolving doubt in his/her epistemological beliefs. There­fore, volition and consciousness are fundamental to understanding the mechanism of change involved in IM and personal epistemological advancement.

Reciprocal causation

Reciprocal causation is considered to be a broader but very important influence in the IM’s mechanism of change (see Figure 4.1). The con­cept of reciprocity expands across many different domains and con­structs in psychology. For example, how people interpret the results of their own behavior informs and changes their environments and the personal factors they possess, thereby informing and altering their sub­sequent beliefs and behaviors (Pajares, 1996). We think that examining some of the psychological roots of reciprocal causation brings clarity to its influence in the IM and is the focus of the following section.

Social cognitive theory. Bandura (1986), in his conceptualization of reciprocal determinism, considers: (1) personal factors (i.e., cognitive, affective, and biological events), (2) behavior, and (3) environmental influences that result in triadic reciprocity. This reciprocity describes the interaction and relations between the three classes of Bandura’s (1986) conception of social cognitive theory. Hence, personal agency is socially rooted and operates within socio­cultural influences, which allows individuals to be viewed as both products and producers of their own environments and social systems (Pajares, 1996). Reciprocity as a concept is an important precursor to understanding reciprocal causa­tion in the IM.

Matthew effect. Another example of a reciprocal effect or reciprocity in the field is the phenomenon called the Matthew effect. There is a verse in the Gospel of Matthew that states, “For to everyone who has shall more be given, and he shall have an abundance; but from the one who does not have, even what he does have shall be taken away” (Matthew 25:29, New American Standard Bible version).

This line has been referred to as the “rich get richer and the poor get poorer” phenomenon (Cunningham and Stanovich, 2003). Reading

The integrative model of personal epistemology108

researchers, for example, have seen time after time that some children arrive at school with a wealthier reading capability than other students who are less fortunate. This has revealed that early reading success is the main contributor to a lifetime of good reading habits and all the benefits that follow. Children who solve the spelling to sound puz­zle early in life operate within a positive feedback loop. This positive feedback loop initiates a reciprocal effect where reading increases the child’s ability to read. The Matthew effect demonstrates that early and efficient reading skill acquisition yields increasing growth rates in read­ing achievement and also in other cognitive skills (Stanovich, 1986; Walberg and Tsai, 1983). Hence, the more children read, the greater their vocabulary and the better their cognitive skills (Cunningham and Stanovich, 2003). We see this kind of effect potentially operating in the IM’s reciprocal causation component. For example, the more students are given a chance to think about and use their beliefs about the nature of knowledge and knowing in the classroom, the greater the chance that they will advance. Additionally, if this is done early in education and continues to be enriched, the possibility of epistemic advancement is all the more heightened.

Reciprocal causation, intelligence quotient (IQ), and social factors. A conceptual introduction to reciprocal causation relating to IQ and social factors also facilitates an understanding of reciprocal causation in the IM. More specifically, Dickens and Flynn (2001) use reciprocal causation to explain the interaction of a person’s phenotypic IQ that influences their environment and the IQ of others with whom they come into contact. If some external factor causes the IQs of some indi­viduals to rise (e.g., better nutrition and health care, parents attending college, technology use) this will improve the environment of others and cause their IQs to rise. Reciprocal causation produces a multiplier effect that inflates both genetic and environmental advantages by a process in which higher IQ leads one into better environments causing still higher IQ, and so on. Harris (1995, 1999) and Harris and Liebert (1991) also discuss reciprocal effects and describe what Dickens and Flynn term multiplier effects as feedback loops and cycles. Dickens and Flynn (2001) call this the social multiplier. This social multiplier, by reacting to the interplay between ability and environment, evolves exponentially. In other words, every rise in individual performance raises the group average, which drives everyone involved to also raise their individual performance to a higher level. This raises the group level to an even higher level and the cycle continues. Even a modest environmental trigger of enhanced performance can become potent

Deanna C. Rule and Lisa D. Bendixen 109

by seizing control of the social multiplier and cause vast performance gains in a relatively short time. In their model of reciprocal causa­tion, Dickens and Flynn (2001) quantify this process and show that initial environmental changes would be enough to explain substantial IQ gains – gains of twenty points over a single generation. These sub­stantial gains in IQ, referred to as the Flynn effect, have been seen in Dutch and Israeli populations between 1952 and 1982 (Flynn, 1987, 1994, 1998).

Conclusion. Similar to Dickens and Flynn, in our discussion of the IM we point out that it is not only an individual’s personal epistemology that influences his or her environment, but also the personal epistem­ologies of others with whom they come into contact. Therefore, if an external factor (e.g., teacher or peer) causes the personal epistemologies of individuals to advance, this might improve the environment of others and will subsequently cause their personal epistemologies to advance. Hence, we have the personal epistemology multiplier (Rule, 2003) (see Figure 4.1). As in the Flynn effect, this personal epistemology multiplier, by reacting with the interplay between personal epistemology and the environment, may also evolve exponentially (see Figure 4.1). In other words, every rise in individual personal epistemology may raise the group average, which might drive everyone involved to advance their individual personal epistemologies. Even a modest trigger of enhanced personal epistemology may possibly become potent by seizing control of the personal epistemology multiplier and cause epistemic gains in a relatively short period of time. Reciprocal causation and the personal epistemology multiplier within the IM sets the stage for researchers to quantify this process and show that initial environmental changes could be enough to explain significant epistemic gains – gains of major importance over a single generation.

Mr. Dyson’s classroom offers an educational example that demon­strates this interplay. Imagine that one of the outcomes of Mr. Dyson’s teaching in his elementary classroom is an advancement in personal epistemological development in the form of students increasing their questioning of theories presented to them. If Mr. Dyson encourages critical thinking and creates an environment where relevant existing theories are carefully evaluated, his students, even at the elementary level, will become more active learners. In doing this the students will eventually encounter epistemic doubt and Mr. Dyson could aid them in resolving that doubt through their own epistemic volition. Hence, the mechanism of change in the IM can be superimposed on the interaction between teachers and students in the classroom. The mechanism can be

The integrative model of personal epistemology110

initiated by either the student or the teacher as long as the environment is conducive to personal epistemological development.

The reciprocal causation component of the integrative model intro­duces the importance of the larger social context to personal epis­temological development. Children, adolescents, and adults need practice in making and defending claims, especially in social contexts where claims must be examined and debated in a framework of alter­natives and evidence. They also need to grow up in a climate where epistemological reasoning is not only evident but accepted, under­stood, valued, and practiced (Bendixen and Rule, 2004). We need to foster individual growth by providing frequent opportunities for the exercise of judgment, but we also need to work toward creating the kind of society in which thinking and judgment are widely regarded as worth the effort they entail (Haerle and Bendixen, 2008; Kuhn and Weinstock, 2002).

Equilibration

As was mentioned previously, the concept of equilibration is fundamen­tal in understanding the adaptive nature of cognition in the context of the IM in general and in its mechanism of change more specifically (see Figure 4.1). For example, an individual may experience equilibration while taking volitional control of his or her doubt. An outcome of this epistemic equilibration is adaptation which is lasting change or lasting personal epistemological development. We see the theories of Piaget (1950, 1985) and Perry (1970) as fundamental to the IM’s considera­tion of equilibration.

Piaget’s equilibration process. As was discussed previously, the under­lying theory for many of the models in personal epistemology rests on the theory and research conducted by Piaget (1950, 1985) and his con­ceptions of cognitive development. Piaget (1970) considers intellectual development as an aspect of genetic epistemology. Genetic epistemology is defined as the study of how knowledge structures grow and develop, with the first rule being collaboration between science and psychol­ogy. This approach engaged developmental psychologists to pursue the juncture of philosophy and psychology as a first step in collaborating (Hofer and Pintrich, 1997). Piaget’s intellectual pursuit was to unveil the nature of knowledge and knowing while particularly addressing the representation, acquisition, development, and mechanism responsible for changing cognitive structures (Flavell et al., 2001). Piaget (1985) proposes that through accommodation (i.e., changing existing schemes to fit new information encountered) and assimilation (i.e., incorporating

Deanna C. Rule and Lisa D. Bendixen 111

new information into existing schemes), the process of equilibration leads to revisions in cognitive systems over time. This pursuit resulted in his model of four stages of cognitive development and its underlying equilibration process.

According to Piaget, our daily responses, whether they are internal or external manifestations of our thought, may take the form of an adapta­tion. Thus, if disequilibrium exists between the environment and the individual, the result may be an action that re­establishes equilibrium, and thus adaptation occurs (Piaget, 1950). Equilibration is a neces­sary factor in mental adaptation. Thus, equilibration is a self­regulatory process leading from one equilibrium point to another equilibrium point (De Lisi and Golbeck, 1999). According to Piaget, the functioning of cognitive systems by the use of accommodation­assimilation engenders a dynamic tension that propels the cognitive system toward revision and change (De Lisi and Golbeck, 1999; Piaget, 1985).

Psychologists other than Piaget discuss equilibrium in their theories as well. For example, Dewey (1933) and Festinger (1957) base their theories on equilibrium and define it “as a principle that affirms a rela­tion between a system (or an organism) and its environment, so that any change in the environment produces an adjustment of the system” (Smith, 2002, p. 1). Hence, the process of equilibration has influenced many theories, including theories of personal epistemology develop­ment (see our previous discussions of King and Kitchener, 1994, and Kuhn and Weinstock, 2002).

In essence, Piaget’s equilibration process enables us to track the adaptive nature of personal epistemological advancement in the IM’s mechanism of change. For instance, we might see epistemic doubt as a more specific form of disequilibrium, and that once resolution strategies are successful equilibrium may be achieved.

Perry’s scheme. We revisit Perry’s (1970) scheme forms of intellectual and ethical development within the context of the IM for a number of reasons. For one, he can be credited for taking aspects of Piaget’s theo ry and situating it more within the area of personal epistemology. His research and model are arguably the most influential within the area of personal epistemology. His theory speaks directly to several pertinent components of our IM, especially the mechanism of change. Finally, his considerations of the educational implications of our work in epis­temological development are still very much alive today and should not be forgotten (see also Moore, 2002).

In the description of his scheme, Perry (1970) considers how people make meaning from their experiences especially when it comes to incon­gruities that are encountered and how they may eventually reach a state

The integrative model of personal epistemology112

of equilibrium. A person also brings with them certain structures or the assumptions and expectations one holds regarding the nature of knowl­edge and knowing. The processes of assimilation and accommodation (in the Piagetian sense) are keys to making sense of these incongruities in experiences. What kind of incongruity it is and its severity will influ­ence how much mental work an individual needs to do to reach some balance between the two processes. Based on his research findings, Perry proposes that assimilation is typically an implicit process where the individual is not cognizant of what is happening. Some accommo­dation processes for the individual, on the other hand, are more explicit and awareness is evident. Additionally, how this may relate to our pre­vious discussion of epistemic volition as needing to be conscious and efficient for lasting change is an interesting question.

Some of Perry’s more general views of development also resonate with the IM. For example, he states that, in essence, growth is synonymous with something that is more desirable than stagnation or regression in development. What is that “more desirable” point in terms of epis­temological development? We address this important question in a later section. In addition, development is not considered by Perry to be a smooth process when moving from position to position in that there are times of stability and times of instability.

What does Perry offer in terms of a mechanism of change in this theo ry? How does this relate to the IM? The impetus for epistemic change in the Perry scheme comes by way of “opportunities” that are presented in the environment and/or from the more internal press/need within the individual. Interestingly (and also in line with Piagetian theory), he points out that what initiated the growth that was encoun­tered by his participants was primarily internal. There may be many “motives” that drive a person through the change process toward matu­ration including “sheer curiosity,” a striving for competence, a need to “make order out of incongruities, dissonances, and anomalies of experience,” and an “urge toward maturation” (Perry, 1970, p. 50–1). Whether or not most of epistemic change within the context of the IM is internally initiated could be investigated by future research.

In addition to development not being a smooth process it is not con­sidered by Perry to be strictly linear either. He speaks of the useful metaphor that exists in development illustrating a tension between two opposing linear forces, one’s awareness of a need for change on the one hand and one’s denial or fear that change may bring “catastrophic disorganization” in one’s beliefs on the other. This tension can be at times very emotional and worrisome for the individual. The stress asso­ciated with some epistemic change has been found in other research

Deanna C. Rule and Lisa D. Bendixen 113

as well (Bendixen, 2002; Chandler et al., 2002). Perry further states that although this more linear metaphor is helpful to a point it does not address the complexity of actual development. “A more proximate metaphor would be drawn from enormously complex self­regulatory systems of nonlinear feedback” (Perry, 1970, p. 52). This more complex and non­linear view of development is consistent with the IM as well and other more recent considerations of epistemological development (e.g., Schommer­Aikins, 2004).

Conclusion. As we have stated, the IM’s mechanism of change and its more general equilibration process have significant roots in the theories of Piaget and Perry. As one advances through the IM’s mechanism of change, one might experience equilibration as a process leading from one equilibrium point to another equilibrium point. There we might also experience a dynamic tension that propels our epistemological beliefs toward revision and change. For example, when a life event triggers doubt in an individual, that individual may start to question what rele­vant truths exist. As that person experiences doubt they might experi­ence disequilibrium. A new point of equilibrium could be achieved as that individual attempts to resolve that doubt. The new point of equi­librium could correspond to a higher level of personal epistemological development or that point of equilibrium can be at the level of personal epistemological growth where everything started. Hence, the individ­ual may have “developed new or better beliefs” or “reaffirmed former beliefs” (Bendixen, 2002, pp. 200–1). Consistent with Piaget and Perry, the IM provides additional explicit detail in the thought processes and emotions associated with being in a state of disequilibrium (i.e., epi­stemic doubt), epistemic volition, and the resulting strategies that may be used to reach a state of equilibrium.

At this point it is important to note that not all shifts in epistemo­logical development necessarily involve large, sweeping, “wholesale” change, and these alterations may not always be accompanied by major emotional upheaval (Chandler et al., 2002; Schraw and Olafson, 2008). We think that the IM can be useful in understanding and researching more of the day­to­day changes a student may experience in the class­room along with more general belief/paradigm shifts one may encoun­ter over longer periods of time.

Summary

As we have shown, the keystones of the IM (i.e., epistemic volition, reciprocal causation, and equilibration) have a rich past and current relevance in educational psychology and psychology. We hope that by

The integrative model of personal epistemology114

examining these keystones more closely we have brought additional clarity and insight to the IM, its mechanism of change, and its potential value for understanding epistemological development. We now turn to a discussion concerning the IM and its role in education.

Educational implications

How can the IM inform teaching and learning in an applied sense? How is it a part of epistemic climate? What would it look like in the classroom? What are our broader educational goals? In this final sec­tion we take the integrative model and its mechanism of change and discuss some of the implications it has for learning and instruction. There is a number of important interconnected elements consistent with the IM that could exist in a classroom that would allow for suc­cessful epistemic advancement of students and we have organized these ideas in Figure 4.2. As can be seen in Figure 4.2, the student and the IM’s mechanism of change (i.e., epistemic doubt, epistemic volition, and resolution strategies) are at the center, linked with cer­tain classroom elements pertaining to them. As we have discussed, as the student moves through the mechanism, their problem­solving behaviors need to be intentional, goal­directed, and metacognitive to be effective. What is also evident in Figure 4.2 is the importance of the teacher in supporting the student as they work through the mech­anism. Teacher support includes setting up incongruities for students, providing compassion and modeling, creating a classroom commu­nity, and utilizing learning materials and procedures effectively. We discuss each of these key elements in more detail in the remainder of the section.

The student

The individual student and their own epistemological progress is obvi­ously a key player in the IM. A lot of what happens developmentally rests on the shoulders of the student and the more they can be made cognizant of this the more potential we see in terms of epistemological advancement.

Awareness. How can students be encouraged to be more aware of their own beliefs and processes of epistemic change? In line with the IM, we argue that increases in awareness are linked with more chances for advancement (see also Muis et al., 2006). Typically, how often do students think about their learning without prompting and, more spe­cifically, how often do they think about aspects of their own epistemo­logical beliefs/development? Based on the literature, we assume that

Deanna C. Rule and Lisa D. Bendixen 115

explicit thinking about epistemological beliefs in the typical classroom is not very typical, but it could be.

Intentionality. We have made several strong statements about the poten­tial power of epistemic volition in the IM’s mechanism of change. How does this translate to students and their work in the classroom? Part of what we are suggesting pertains to the need to convince students that they are, indeed, in charge of their own epistemic views, and this is not an easy task by any means. In other words, this allows students to play a bigger part in their own learning (Perry, 1970). As with most areas of learning, for students to become more intentional they need practice at doing it in a variety of settings and over a long period of time. As students experience the inevitable setbacks and, more importantly, successes in becoming more intentional, they will be more likely to value the role of epistemic volition in their own learning. The educational benefits of epistemic volition must be there for students as well. If Mr. Dyson, for example, did not value students taking charge of their own beliefs and

Classroomcommunity

Learningmaterial andprocedures

Compassion/modeling

Goal: epistemic advancement

Student

1 Epistemic doubt2 Epistemic volition3 Resolution strategies

Problem-solving

Goal-directedlearning

Intentionality

Awareness

Teacher support

Setting upincongruities

Thank U 2 FN 4 TCB

Figure 4.2: Key classroom elements for epistemic advancement

The integrative model of personal epistemology116

learning, he would not have bothered devoting a significant amount of his classroom time and teaching energy encouraging his students to experience not only epistemic doubt but the chance to resolve it through volition (e.g., students examining their own beliefs and making their own judgments about the merits of different problem­solving approaches).

Goal-Directed. Is it common for students to be conscious of the short­ and long­term goals? Consistent with the literature on goal theory (e.g., Ames, 1992; Harackiewicz et al., 2002), the more students are planful and goal­directed in their own learning, the more likely they are to be successful in their achievement. We would also argue that being goal­directed in learning allows students to be more aware of the role of their own epistemological beliefs. Mr. Dyson, for example, communicated to his students that he was not after the one right way to solve the math­ematics problem but wanted the students to find different ways and to judge them together. This kind of goal and others like it carries a power­ful message to students in terms of the parameters of their own learn­ing and beliefs. Making learning goals more apparent and accessible to students is a valuable endeavor in the classroom. In addition, allowing students to think about and express their own goals and expectations for learning (e.g., “I just want to find the one right theory” vs. “I need to make some judgments about which theory has more merit”) will influ­ence their behavior in the classroom. This would be another step in making the epistemic change process more transparent and successful.

Problem-solving. We think that an important element of epistemic advancement in the classroom pertains to the kinds of problem students are being asked to solve. The important role of ill­defined problems (i.e., problems that do not have one right answer and/or one correct way of solving them) and epistemological beliefs has been documented and discussed (Muis et al., 2006; Schraw et al., 1995). Problems such as these that are more complex in nature are more conducive to stu­dents thinking about their own epistemological beliefs. We would also argue, however, that this does not preclude the value of well­defined problems in that they could also be closely examined epistemically by students and teachers as well. For example, it is often important to rely on authority in certain problem­solving situations (i.e., the objective aspects of knowledge and knowing). In general, problems that require critical thinking and argumentation are crucial for epistemic advance­ment to occur (Kuhn and Weinstock, 2002; Schraw, 2001).

Teacher support

In our consideration of the epistemic climate of the classroom, the teacher is central in the epistemological development of students (Bendixen and

Deanna C. Rule and Lisa D. Bendixen 117

Rule, 2004; Haerle and Bendixen, 2008) (see Figure 4.2). Consistent with Perry (1970) we see teachers as a vital source of student support through the process of epistemic change. What do we mean by support? Our interpretation of teacher support includes: (1) setting up incon­gruities for students, (2) modeling and compassion, (3) building class­room community, and (4) using learning materials and instructional procedures effectively.

Setting up incongruities. In discussing the implications of his scheme, Perry (1970) recommends that teachers provide “calculated incongru­ities” for their students and that these are a requirement for epistemic movement to occur. We see this as consistent with the epistemic doubt component, in other words, creating doubt with support. There are many ways that students’ beliefs could be challenged in the everyday classroom (e.g., working with real­world conflicting data, setting up classroom debates, etc.) (Hofer, 2001; Kardash and Scholes, 1996). According to Perry (1970), however, a teacher must first be cognizant of where their students are in terms of their views of knowledge and knowing. This is an interesting question for the field. Currently, are there adequate materials available for teachers to assess their students’ epistemological beliefs/development (Haerle, 2006; Muis et al., 2006)? How specific do students beliefs need to be measured to be helpful? We would argue that many more of the methodological issues in the field need to be addressed before they are passed on to teachers. In addition, it is not enough to know where students’ beliefs are and then challenge them. A teacher should be prepared to address and plan for the epis­temic volition and resolution strategies components discussed in the IM’s mechanism as well.

Compassion/modeling. Calls for the need to understand and address the affective side of epistemological development are becoming more and more apparent (e.g., Pintrich, 2002). Perry (1970) was certainly one of the first to give a glimpse of and discuss at length the more emotional sides of development. As has been pointed out, epistemic doubt can at times be difficult, risky, and lonely for students (e.g., Bendixen, 2002; Chandler et al., 1990). Compassion for their students along these lines would be part of the support a teacher could pro­vide. Allowing students to discuss and journal their feelings would be examples of ways to acknowledge and support this aspect of epistemic development.

Teachers offering more “visibility in their own thinking” or mod­eling their own epistemic thoughts and challenges could also be a powerful tool for support (Perry, 1970, p. 213; see also Schraw, 2001). Would students be more willing to accept the challenges of epistemic change if their teacher modeled this as an important part of learning?

The integrative model of personal epistemology118

For example, Mr. Dyson could “think aloud” along with his students as they work their way through different ways the mathematics problems could be solved, including posing some of his own pertinent epistemic questions that he may not have the answers for. The key here would be to make the components of the mechanism of change process more metacognitive or transparent for students to help them understand and be encouraged by it.

Classroom community. As students experience their own epistemic doubt they may feel as if they are alone and uneasy in these questions and concerns (Bendixen, 2002; Chandler et al., 1990). These feelings can certainly be minimized if students perceive they are a part of a trust­ing classroom community and that many of their fellow students (and their teacher at times) are in the same boat (Perry, 1970). Of course, the teacher can set this up in a number of ways, such as involving students more in the workings of the classroom and conveying to the students that they too have experienced, for example, epistemic doubt in their learning.

Learning materials and procedures. Examining the epistemic aspects of learning materials (e.g., textbooks, curriculum, etc.) and procedures (e.g., pedagogy, assessment, etc.) that teachers use is also an impor­tant form of support (Haerle, 2006). For example, exposing students to textbooks and curriculum that is not so absolute in appearance could help students become accustomed to the relative nature of some knowledge (Muis et al., 2006; Wineburg, 1991). In terms of assessment, when is correcting important and when is supporting the process more important is a noteworthy question (Perry, 1970). Teacher feedback supporting the process of learning in lieu of always focusing on the end product is consistent with epistemic advancement. Additionally, how often are students asked to create and evaluate and be assessed in doing these things? Finally, more constructivist teaching practices would also be more supportive of the epistemic change process (e.g., Muis et al., 2006).

What are our educational goals?

Our final thoughts pertain to the more general educational goals we may have in terms of the IM and its portrayal of epistemological development. At this point we ask where does all of this lead us? What are we developing toward? As we stated in an earlier section, Perry (1970), in general, viewed development and growth as synonymous with something that is more desirable than stagnation or regression. What

Deanna C. Rule and Lisa D. Bendixen 119

is that “more desirable” point in epistemological development? Is it or should it be our explicit educational goal? What have others labeled it in our field?

In the first pages of this chapter we briefly reviewed some of the research literature in personal epistemology that provides empir­ical evidence that more advanced epistemic beliefs are directly and indirectly associated with increased student achievement. We think it would be extremely helpful to further clarify for students, educators, and researchers what those more advanced epistemic beliefs would be and how they can be translated into effective educational goals. The labels for these more advanced points of epistemological devel­opment include, for example, commitments within relativism (Perry’s scheme (1970)), reflective judgment (King and Kitchener, 1994), and evaluativism (Kuhn and Weinstock, 2002). In the IM and the class­room elements that support it the goal is epistemic advancement in students toward evaluativist thinking about knowledge and knowing. What this integration of objective and subjective knowing means and looks like in the classroom needs to be further elucidated in future work in the field. If our educational goals and evaluativism are clearer we see a number of positive outcomes. A major benefit of evaluativsm includes students becoming better prepared to deal with the compli­cated and relativistic world in which they live (Haerle and Bendixen, 2008; Kuhn and Weinstock, 2002; Perry, 1970). Finally, as we have stated, we do see a fair amount of the burden resting on the educa­tors of students in this regard. In other words, do the teacher and the broader educational community have a responsibility to promote epis­temic growth and development and to support students as they navi­gate through this process (Perry, 1970)? We respond to this question with a resounding yes.

R EF ER ENCES

Ames, C. (1992). Classrooms: Goals, structures and student motivation. Jour-nal of Educational Psychology, 84, 261–71.

Anderson, R. C., Reynolds, R. E., Schallert, D. L., and Goetz, E. T. (1977). Frameworks for comprehending discourse. American Educational Research Journal, 14, 367–81.

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall.

Baxter Magolda, M. B. (1992). Knowing and reasoning in college: Gender-related patterns in students’ intellectual development. San Francisco: Jossey­Bass.

(2001). Making their own way: Narratives for transforming higher education to promote self-development. Sterling, VA: Stylus.

The integrative model of personal epistemology120

(2004). A constructivist conceptualization of epistemological reflection. Educational Psychologist, 39, 31–42.

Belenky, M. F., Clinchy, B. M., Goldberger, N. R., and Tarule, J. M. (1986). Women’s ways of knowing: The development of self, voice, and mind. New York: Basic Books.

Bendixen, L. D. (2002). A process model of epistemic belief change. In B. K. Hofer and P.R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

Bendixen, L. D. and Rule, D. C. (2004). An integrative approach to personal epistemology: A guiding model. Educational Psychologist, 39, 69–80.

Burr, J. E. and Hofer, B. K. (2002). Personal epistemology and theory of mind: Deciphering young children’s beliefs about knowledge and know­ing. New Ideas in Psychology, 20, 199–224.

Chandler, M. J., Boyes, M., and Ball. L. (1990). Relativism and stations of epi­stemic doubt. Journal of Experimental Child Psychology, 50, 370–95.

Chandler, M. J., Hallett, D., and Sokol, B. W. (2002). Competing claims about competing knowledge claims. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (154–68). Mahwah, NJ: Lawrence Erlbaum Associates.

Corno, L. (1993). The best­laid plans: Modern conceptions of volition and educational research. Educational Researcher, 22(2), 14–22.

Cunningham, A. E. and Stanovich, K. E. (2003). Reading can make you smarter: The more children read, the greater their vocabulary and the better their cognitive skills [Electronic version]. Principal – what principals need to know about reading, 83(2), 34–9.

De Lisi, R. and Golbeck, S. L. (1999). Implications of Piagetian theory for peer learning. In A. M. O’Donnell and A. King (Eds.), Cognitive perspectives on peer learning. Mahwah, NJ: Lawrence Erlbaum Associates.

Dewey, J. (1933). How we think. New York: D. D. Heath.Dickens, W. T. and Flynn, J. R. (2001). Heritability estimates versus large

environmental effects: The IQ paradox resolved. Psychological Review, 108, 346–69.

Flavell, J. H., Miller, P. H., and Miller, S. A. (2001). Cognitive Development (4th edn.). Upper Saddle River, NJ: Prentice Hall.

Festinger, L. A. (1957). Theory of cognitive dissonance. Stanford, CT: Stanford University Press.

Flynn, J. R. (1987). Massive gains in 14 nations: What IQ tests really measure. Psychological Bulletin, 101, 171–91.

(1994). IQ gains over time. In R. J. Sternberg (Ed.), Encyclopedia of human intelligence (617–23). New York: Macmillan.

(1998). Israeli military IQ tests: Gender differences small; IQ gains large. Journal of Biosocial Science, 30, 541–53.

Haerle, F. C. (2006). Personal epistemologies of fourth-graders: Their beliefs about knowledge and knowing. Oldenburg, Germany: Didaktisches Zentrum.

Haerle, F. C. and Bendixen, L. D. (2008). Personal epistemology in elementary classrooms: A conceptual comparison of Germany and the United States

Deanna C. Rule and Lisa D. Bendixen 121

and a guide for future cross­cultural research. In M. S. Khine (Ed.), Knowing, knowledge and beliefs: Epistemological studies across diverse cultures (151–76). New York, NY: Springer.

Hammer, D. and Elby, A. (2002). On the form of a personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

Harackiewicz, J. M., Barron, K. E., Pintrich, P. R., Elliot, A. J., and Thrash, T. M. (2002). Revision of achievement goal theory: Necessary and illu­minating. Journal of Educational Psychology, 94(3), 638–45.

Harris, J. R. (1995). Where is the child’s environment? A group socialization theory of development. Psychological Review, 102, 458–89.

(1999). The nurture assumption. New York: Simon and Schuster.Harris, J. R. and Liebert, R. M. (1991). The child: A contemporary view of devel-

opment (3rd edn.). Englewood Cliffs, NJ: Prentice Hall.Hofer, B. K. (2001). Personal epistemology research: Implications for learning

and teaching. Educational Psychology Review, 13, 353–83. (2004). Epistemological understanding as a metacognitive process: Think­

ing aloud during online searching. Educational Psychologist, 39, 43–56.Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological

theories: Beliefs about knowledge and knowing and their relation to learn­ing. Review of Educational Research, 67, 88–140.

(2002). Personal epistemology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

Kahneman, D. (2003). A perspective on judgment and choice. American Psych-ologist, 58, 697–720.

Kardash, C. M. and Scholes, R. J. (1996). Effects of preexisting beliefs, epis­temological beliefs, and need for cognition on interpretation of controver­sial issues. Journal of Educational Psychology, 88, 260–71.

King, P. and Kitchener, K. S. (1994). Developing reflective judgment: Under-standing and promoting intellectual growth and critical thinking in adolescents and adults. San Francisco: Jossey­Bass.

Kuhn, D. (1991). The skills of argument. New York: Cambridge University Press.

Kuhn, D. and Weinstock, M. (2002). What is epistemological thinking and why does it matter? In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

Mason, L. (2002). Developing epistemological thinking to foster conceptual changes in different domains. In M. Limon and L. Mason (Eds.), Recon-sidering conceptual change: Issues in theory and practice (301–35). Dordrecht, The Netherlands: Kluwer.

(2003). Personal epistemologies and intentional conceptual change. In G. M. Sinatra and P. R. Pintrich (Eds.), Intentional conceptual change (199–236). Mahwah, NJ: Lawrence Erlbaum Associates.

Moore, W. S. (2002). Understanding learning in a postmodern world: Recon­sidering the Perry scheme of intellectual and ethical development. In

The integrative model of personal epistemology122

B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychol-ogy of beliefs about knowledge and knowing (17–36). Mahwah, NJ: Lawrence Erlbaum Associates.

Muis, K. R., Bendixen, L. D., and Haerle, F. C. (2006). Domain­generality and domain­specificity in personal epistemology research: Philosophical and empirical reflections in the development of a theoretical framework. Educational Psychology Review, 18(1), 3–54.

Pajares, F. (1996). Self efficacy beliefs in academic settings. Review of Educa-tional Research, 66, 543–78.

Perry, W. G. (1970). Forms of intellectual development and ethical development in the college years: A scheme. New York: Holt, Rinehart and Winston.

Piaget, J. (1950). The psychology of intelligence. New York: Routledge. (1970). Psychology and epistemology. New York: Viking Press. (1985). The equilibration of cognitive structures: The central problem of intellectual

development. Chicago: University of Chicago Press.Pintrich, P. R. (2002). Future challenges and directions for theory and research

on personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Per-sonal epistemology: The psychology of beliefs about knowledge and knowing (389–414). Mahwah, NJ: Lawrence Erlbaum Associates.

Qian, G. and Alverman, D. (2000). Relationship between epistemological beliefs and conceptual change learning. Reading and Writing Quarterly, 16, 59–74.

Reynolds, R. E. and Sinatra, G. M. (2009). The cognitive revolution in scientific psychology: Epistemological roots and impact on reading research. In press.

Reynolds, R. E., Sinatra, G., and Jetton, J. (1996). Views of knowledge acquisi­tion and representation: A continuum from experience­centered to mind­centered. Educational Psychologist, 31(2), 93–104.

Rule, D. C. (2003). The economics of personal epistemology. Unpublished manu­script, University of Nevada, Las Vegas.

Schommer­Aikins, M. (2004). A systemic approach to the conceptualization and study of epistemological beliefs: When researchers coordinate and cooperate. Educational Psychologist, 39, 19–30.

Schraw, G. (2001). Current themes and future directions in epistemological research: A commentary. Educational Psychology Review, 13, 451–65.

Schraw, G., Dunkle, M. E., and Bendixen, L. D. (1995). Cognitive processes in well­defined and ill­defined problem solving. Applied Cognitive Psychol-ogy, 9, 523–8.

Schraw, G. J. and Olafson, L. J. (2008). Assessing teachers’ epistemological and ontological worldviews. In M. S. Khine (Ed.), Knowing, knowledge and beliefs: Epistemological studies across diverse cultures (25–44). New York, NY: Springer.

Sinatra, G. M. and Pintrich, P. R. (2003). The role of intentions in conceptual change learning. In G. M. Sinatra and P. R. Pintrich (Eds.), Intentional conceptual change. Mahwah, NJ: Lawrence Erlbaum Associates.

Smith, L. (2002). Critical readings on Piaget. New York: Routledge.

Deanna C. Rule and Lisa D. Bendixen 123

Stanovich, K. E. (1986). Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy. Reading Research Quar-terly, 21, 360–407.

Stanovich, K. E. and West, R. F. (2000). Individual differences in reason­ing: Implications for the rationality debate. Behavioral and Brain Sciences, 23, 645–65.

Vosniadou, S. (2003). Exploring the relationships between conceptual change and intentional learning. In G. M. Sinatra and P. R. Pintrich (Eds.), Inten-tional conceptual change (377–406). Mahwah, NJ: Lawrence Erlbaum Associates.

Walberg, H. J. and Tsai, S. (1983). Matthew effects in education. American Education Research Journal, 20, 359–73.

Wineburg, S. (1991). Historical problem solving: A study of the cognitive proc­esses used in the evaluation of documentary and pictorial evidence. Jour-nal of Educational Psychology, 83, 73–87.

124

5 An epistemic framework for scientific reasoning in informal contexts

Fang­Ying Yang National Taiwan Normal University, Taipei, Taiwan

Chin­Chung Tsai National Taiwan University of Science and Technology, Taipei, Taiwan

Introduction

In the science education community, there is a growing consensus that in addition to conceptual knowledge, we need to introduce learners to another important facet of science, that is, how we create new knowledge. In other words, students should be better educated in the use of certain established ways of thinking in science (e.g., Duschl, 1990; Lawson et al., 2000). The “certain established ways of thinking in science” are commonly referred to as scientific reasoning, which is portrayed by philosophers of science as a process of argumentation (Giere, 1991; Seigel, 1988; Toulmin, 1958), because it involves the evaluation of evidence to support a theory or claim. In schools, scientific reasoning is usually presented in domain­specific contexts such as physics, chemistry, life sciences, and so forth. Neverthe­less, as Seigel (1988) has noted, the commitment to evidence is an impera­tive trait of rational reasoning in many disciplines, although the form it takes may vary with the disciplines. Even in everyday situations, testing of the possibilities with accountable evidence or reasons and searching for possibilities are critical for decision­making (Baron 1988; Kuhn, 1991; Lawson et al., 2000; Perkins and Salman, 1989). Hence, although scien­tific reasoning is often discussed within specific knowledge domains, as Kuhn (1993) pointed out, it represents a domain­independent mode of argumentative reasoning.

The development of scientific reasoning has been widely discussed in psychological research (Zimmerman, 2000). Nevertheless, most studies are placed in domain­specific contexts and deal with well­structured problems with only a few exceptions (Kuhn, 1991, 1993). Although there are quite a few studies examining factors contribut­ing to the performance of scientific reasoning, the major focus is the role of domain­specific knowledge (Zimmerman, 2000). Only recently

Fang­Ying Yang and Chin­Chung Tsai 125

have psychologists started to raise issues concerning the effects of other underlying mediators, particularly personal epistemic beliefs (Duschl and Osborn, 2002; Hofer and Pintrich, 1997; King and Kitchener, 1994; Pintrich, 2002; Sandoval, 2005). However, in the area of science education, the influence of epistemic beliefs has not yet been addressed enough (Kelly and Duschl, 2002). In this chapter, we attempt to develop a theoretical framework based on empirical data obtained from chil­dren and teenagers, linking the performance of scientific reasoning in informal contexts with the development of personal epistemology.

Scientific reasoning in informal contexts

As emphasized by philosophers of science, psychologists, and science educators, science is characterized by the processes of hypothesis/theory generalization (process of discovery) and hypothesis testing (process of justification) (e.g., Duschl, 1990; Kuhn, 1970; Newton et al., 1999; Toulmin, 1958; Zimmerman, 2000). It is through the process of argu­ment in which scientific conjectures are justified by accountable evi­dence, that scientific knowledge becomes public (Giere, 1991; Newton et al., 1999). According to Toulmin (1958), the structure of argument consists of four basic components, including data (evidence), claim (theory), warrants, and backing. What counts as theory and evidence in science is socially agreed upon by the scientific community. In short, the rationality of science is founded on the ability to construct per­suasive and convincing arguments that relate explanatory theories to observational data (Duschl and Osborne, 2002).

As mentioned previously, scientific reasoning is often introduced in schools in domain­specific contexts in which well­structured problems with standardized answers are usually employed as the major problem context. However, scientific reasoning habits are important in daily expe­riences because they provide important ways to make rational and sound judgments about controversial issues in social contexts. Examples of these issues include the use of nuclear power, the development of genetic tech­nology, and the reduction of the global warming effect that increasingly have significant consequences for future society. Kuhn (1993), while rec­ognizing science as argument, has suggested that to describe scientific reasoning as argument is also sensible in everyday situations.

According to Perkins (1985), reasoning over well­structured prob­lems which have fixed premises and a well­formed argument that leads to a final conclusion is defined as formal reasoning. On the other hand, the reasoning process that is more directly applicable to situations where the problem is ill­structured and requires the generation of an

An epistemic framework for scientific reasoning126

argument to support a claim based on evidence relevant to the problem is defined as informal reasoning (Means and Voss, 1996; Voss et al., 1991; Willis and Schaie, 1993). Perkins (1985) argues that reasoning in informal contexts is a process of situation­modeling which requires the reasoner to build a model of a situation that articulates the dimensions and factors involved in the issues. Such a model usually involves one or more imagined scenarios. It will evoke a variety of common sense, causal, and intentional principles to construct and weigh the plausibil­ity of alternative scenarios (Perkins, 1991). In short, formal reasoning and informal reasoning share the same form of argumentation, that is, generating an argument based on accountable evidence, they differ largely in the nature of problems to be solved, the bases of argument, and criteria for justification.

As mentioned previously, scientific reasoning is described as the processes of hypothesis/theory generalization (process of discovery) and hypothesis testing (process of justification). When such a reason­ing process is applied to everyday situations, it falls into the category of informal reasoning. To highlight the importance of scientific reasoning in everyday situations, we use scientific reasoning in informal contexts to indicate informal reasoning about real­life issues that involve debates or discussions over uncertain scientific information.

Although the importance of scientific reasoning habits is recognized, Kuhn (1991) reported that adult thinkers often do not have adequate abilities to make scientific arguments regarding social issues. Many other studies also found similar results regarding social issues that con­cern the implementation of science and technology (Jiménez­Aleixandre and Pereiro­Munoz, 2002; Kortland, 1996; Perkins, 1989; Yang and Anderson, 2003; Yang, 2004). One of the challenges presented to science educators, and maybe to all educators, therefore, is how to create learning situations where these powerful forms of thought can be anchored in situ­ations that may be generalizable beyond science classrooms. To provide this kind of science instruction, a deeper understanding concerning stu­dents’ difficulty of practicing scientific reasoning in informal contexts is needed.

Scientific reasoning and personal epistemology

In psychological and educational research, many studies have identified that personal epistemological beliefs regarding the nature of knowl­edge, knowing, and/or learning mediate thinking, decision­making, and learning approaches (e.g., Hofer and Pintrich, 1997; Hofer and Pintrich, 2002; Perry, 1970; Schommer­Aikins, 1993; Tsai, 1998, 2000, 2004). The personal epistemological beliefs usually referred to as personal

Fang­Ying Yang and Chin­Chung Tsai 127

epistemology are considered to be the highest level of cognition that regulates cognitive processes and development (Kitchener, 1983). In the literature, there are various theoretical models about personal epis­temology. For example, many researchers emphasize the developmental nature (e.g., Baxter Magolda, 1992; King and Kitchener, 1994; Kuhn, 1999; Perry, 1970); some claim the independence between different epistemological dimensions (Schommer­Aikins, 2002), whereas others argue for the systematic or ecological interrelation among an identifi­able set of dimensions of beliefs (e.g., Hofer, 2001).

In our view, it might be needless to argue about which model can best represent one’s personal epistemology. Different models with different theoretical bases allow researchers to discuss diverse relations between epistemic beliefs and cognitive processes. Some of the relations undoubt­edly are influenced by cognitive development. Based on previous research results (Yang, 2004, 2005; Yang et al., 2005), which will be discussed later, developmental perspectives were employed in this study.

The original developmental model was proposed by Perry (1970), who suggested that the forms of personal epistemology are a result of educa­tional experiences and they progress through stages including dualism, multiplicity, contextual relativism, and commitment within relativism. Individuals in different developmental stages or positions exhibit dif­ferent contemplations about knowledge and learning. Table 5.1 shows the content for “view of knowledge” in Perry’s scheme (Perry, 1970) with respect to different epistemological positions. Descriptions for the other views about learning can be found in Perry’s book (1970), Forms of intellectual and ethical development in the college years.

Other developmental models, although using different terminolo­gies, also point to a similar progression. For instance, Kuhn (1999) found a person’s epistemological understanding evolving from realist, absolutist, and multiplist to the evaluativist form of thinking. Briefly, the realist thinks that knowledge is certain and comes from external sources while assertions are copies of external reality. The absolutist also believes that knowledge is certain but sees assertions as facts that are correct or incorrect in their representation of reality. The multiplist thinks that knowledge is generated by human minds and is uncertain while assertions are opinions. For the evaluativist, knowledge is uncer­tain but assertions can be evaluated and compared according to criteria of argument and evidence. In short, as Hofer (2001) pointed out, per­sonal epistemology from the developmental point of view evolves from an absolutist or dualist view of knowledge, followed by a multiplicity standpoint that allows for the existence of different viewpoints, to the evaluativist stance that knowledge is actively constructed by the knower and is justifiable.

An epistemic framework for scientific reasoning128

The associations between the development of personal epistemology and reasoning have been documented in many studies. King and Kitch­ener (1994) proposed the reflective judgment model which depicts that development of reflective thinking (that is, the evaluation and integra­tion of existing data and theory into a solution about the problem at hand) is accompanied with the development of epistemic assumptions about knowledge and justification. In their model, reflective reasoning progresses from pre­reflective to quasi­reflective, to reflective thinking. Similarly, Kuhn and others (Kuhn, 1991, 1999; Kuhn and Weinstock, 2002; Mason and Scirica, 2006) have demonstrated the developmental association between epistemological understanding and argumentative reasoning. As described later, our studies also show that performance of scientific reasoning on science­based, ill­structured problems is cor­related with personal epistemological beliefs (Yang, 2004, 2005).

Table 5.1. Descriptions for the “view of knowledge” with respect to different epistemological positions in the Perry scheme1

Epistemological position View of knowledge

Dualism (position 2) All knowledge is known. There is a certainty that right and wrong answers

exist for everything. Knowledge is a collection of information.

Early multiplicity (position 3) Most knowledge is known. All is knowable. Certain that there exists a right way to find the

right answers. Realization that some knowledge domains are

fuzzy.

Late multiplicity (position 4) In some areas we still have certainty about knowledge.

In most areas we really don’t know anything for sure.

Certainty that there is no certainty.

Contextual relativism (position 5)

All knowledge is contextual. All knowledge is disconnected from any concept

of absolute truth. However, right and wrong, adequate and

inadequate, appropriate and inappropriate, can exist within a specific context.

1 Adapted from Perry’s book (1970), Forms of intellectual and ethical development in the college years.

Fang­Ying Yang and Chin­Chung Tsai 129

Although many studies have suggested that the performance of sci­entific reasoning in informal contexts can be explained by the develop­ment of personal epistemology, most relevant studies were conducted with adult or adolescent subjects. We lack sufficient information about children’s reasoning behavior and epistemological development. In psychological research about theory of mind, two important find­ings provide clues about the origin of personal epistemology. First, the development of the ability to reason and act in accordance with beliefs starts in early childhood (Lee and Homer, 1999; Wellman et al., 2001). Second, the ability to distinguish objective reality and sub­jective belief, which is vital to the understanding of evidence evolves with age (Astington et al., 2002; Burr and Hofer, 2002; Mansfield and Clinchy, 2002). These two findings suggest that the development of epistemological thinking should start at a young age (Hofer, 2000) and should mediate the development of scientific reasoning that has been found to start at an early age as well (Zimmerman, 2000). However, until more is known about the relation between scientific reasoning and personal epistemological beliefs in young thinkers, our conclusion about the developmental link is only suppositional.

Previous studies on scientific reasoning in informal contexts

Based on previous research results (Yang and Anderson, 2003; Yang, 2004; Yang, 2005; Yang et al., 2005) and with some newly collected data, we hope to construct a theoretical framework that can link the performance of scientific reasoning in informal contexts to the devel­opment of personal epistemology. The modes of scientific reasoning discussed here are targeted on the process of hypothesis testing and include the coordination of theory and evidence and reflective think­ing. According to Kuhn (1991), the two processes of reasoning when being applied to informal contexts reflect major characteristics of argumenattive reasoning in everyday thoughts. To be brief, we want to explore if an individual thinker is capable of justifying claims or opin­ions based on plausible evidence, and whether the person is aware of his/her own reasoning process. The target issues discussed in this chap­ter are science­related, in­dispute issues (one is a socio­scientific issue, the other pure­science issue) that are ill­structured by nature due to the uncertainty of expert information involved for discussions. To begin with, we first present several previous studies about scientific reasoning in informal contexts with sixth­ and tenth­grade subjects.

An epistemic framework for scientific reasoning130

Study 1: Senior high school students’ preference and reasoning modes about nuclear energy use (Yang and Anderson, 2003)

Purpose. The purpose of the study was to explore high school students’ modes of reasoning and intention in using scientific information in evaluating the issue about use of nuclear energy. Although Study 1 did not specifically probe personal epistemology, the result of the study revealed the need for research on the relation between scientific rea­soning and personal epistemology.

Participants. The participants were 182 twelfth­graders from four intact classes at two academic senior high schools in a mid­size city in Taiwan. There were eighty­six male and ninety­six female students.

Measure. An information preference questionnaire with a five­point Likert scale was developed to assess students’ cognitive orientation toward scientific or social information relevant to the use of nuclear energy. According to the preference scores, participants were grouped into three preference groups, namely scientifically oriented, socially oriented, and equally disposed information groups. About ten of them in each group were selected for in­depth interviews to uncover reasoning modes and their intention to use scientific information in reasoning.

Results. It was found that when making decisions on the use of nuclear energy, adolescent thinkers displayed different cognitive preferences toward scientific and social information, and they used social and sci­entific information in different ways to justify ideas and/or opinions. Basically, those who regarded scientific information as relatively more important than social information (the scientifically oriented groups) tended to be more willing to take into account scientific information in making judgments. Those who emphasized the importance of social information (the socially oriented group) preferentially used social infor­mation. Interestingly, it was those who took into account equally the science and social information (the equally disposed group) produced better decision­making results. While scientific information played a role in supporting personal theories, social information was frequently used to justify scientific information. Although it was not explicitly dis­cussed in the research report, the cognitive preferences toward different types of information during decision­making were apparently related to thinkers’ beliefs about the source of knowledge. The different functions of scientific and social information that appeared in reasoning seemed to indicate thinkers’ ways of knowing. Based on the results of the study, we postulated that the different modes of reasoning indicated underly­ing effects of personal epistemological beliefs.

Fang­Ying Yang and Chin­Chung Tsai 131

Study 2: Exploring high school students’ use of theory and evidence in an everyday context: the role of scientific thinking in environmental science decision-making (Yang, 2004)

Purpose. The previous study explored the general modes of reasoning and uses of social and scientific information in decision­making on a socio­scientific issue. To specify reasoning competence, this study examined students’ use of some particular scientific reasoning skills, including the coordination of theory and evidence and reflective rea­soning in thinking about a social issue that involved the implication of science and technology. In addition, factors that might contribute to reasoning performance, such as conceptual knowledge and gender effect, were also studied.

Participants. Forty­five males and forty­five females in the tenth­grade from two intact classes in a mid­size academic high school in Taiwan participated in the study.

Measure. An open­ended­question survey was designed for data col­lection. The target issue concerned some town residents’ resistance to a well­drilling project planned by a water company. The story of the issue and interview questions are listed in Appendix A. Questions were pos­ited for students to justify the event. In addition, students were asked to reflect on their own ideas and provide suggestions for making bet­ter judgments. Before the study, a supplemented instruction concern­ing the use of underground water was presented to the participants for the purpose of providing relevant science concepts. Assessed by the flow­map method, which qualitatively examined learners’ knowledge structure (Anderson and Demetrius, 1993) regarding the target issue, a significant gain in conceptual knowledge was found after the supple­mented instruction.

Results. The study showed that 70 percent of participants were able to formulate personal theories regarding the cause of land subsidence based on available information and then make inferences based on their theories. Three kinds of personal theories were identified. One was that the previous earthquake caused the land subsidence, another was that the excessive use of underground water resulted in land subsidence, and the other was that the combination of the former two cases con­tributed to the subsidence. Most subjects attended to either the first or second hypothesis rather than the last one which had more scientific merit. When asked to justify the claim given by the water company, only 57 percent of the subjects were able to identify the role of evidence to support theories. Further Chi­square analysis showed no association between the use of personal theory and the identification of the role of

An epistemic framework for scientific reasoning132

evidence in reaching a conclusion, indicating a difficulty in the coordi­nation between theory and evidence. Explored by one­way analysis of variance (ANOVA), it was found that concept achievement was associ­ated with the use of personal theory, but no association was found with the use of evidence. Seemingly, the ability to identify evidence in reason­ing was not necessarily related to a thinker’s background knowledge.

The analysis of students’ self­reflections showed that most subjects (86 percent) were not confident about their thoughts, and the majority tended to attribute their uncertainty to insufficient information pro­vided to them and insufficient knowledge about the scientific issue. In addition, some students tended to think that the words or perspec­tives presented by experts guaranteed their validity. Only a few sub­jects stated that there might be alternatives, or that there is no absolute right or wrong about the issue. It seemed that most students believed that knowledge can provide certain answers to all problems even if the answers are not immediately attainable. From an epistemological point of view, such a view toward knowledge indicates that the students had not developed a more advanced form of personal epistemology.

In light of the epistemological terminology, most students in the study were identified as absolutists or early multiplists at best who believe all things are knowable, but are starting to realize that some knowledge domains are fuzzy (Kuhn, 1999; Perry, 1970). For most participants, since all is knowable, information provided by the water company should be right or wrong and experts were those who could tell the cor­rectness of information. In this case, the important role of evidence was easily ignored. In short, students’ self reflections revealed that they held rather simple forms of personal epistemology which explained the poor performance of the coordination of theory and evidence in the context of socio­scientific issues.

Study 3: Student views concerning evidence and the expert in reasoning a socio-scientific issue and personal epistemology (Yang, 2005)

Purpose. As Hofer and Pintrich (1997) pointed out, personal epistemol­ogy is usually regarded as a domain­general construct. However, there is an increasing number of studies showing that while general personal epistemological beliefs do exist, there are also context­ and/or domain­dependent beliefs that are correlated with domain­general beliefs (e.g., Hofer, 2000; Louca, et al., 2004; Muis et al., 2006; Schommer­Aikins, 2005). The purpose of the study was thus to examine if an individual’s epistemological perspective about science is aligned with the general construct of personal epistemology.

Fang­Ying Yang and Chin­Chung Tsai 133

Participants. Participants of the study were seventy tenth­grade stu­dents who came from two intact classes in an academically oriented high school in Taiwan.

Measure. Students’ views toward scientific evidence and expert opin­ions which reveal epistemological perspectives about the nature of sci­ence were assessed by an open­ended questionnaire (see Appendix B). The epistemological perspectives of science were then cross­ compared with respondents’ personal epistemology assessed by the quantitative instrument: learning environment preference (LEP) questionnaire (Moore, 1989), which was designed based on the Perry scheme (1970). The LEP questionnaire consists of five sections in accordance with the five domains of the learning environment identified in Perry’s study (1970). Each section contains thirteen items describing the characteris­tics of the corresponding domain with respect to epistemological posi­tions two to five, indicating dualist, early multiplicity, late multiplicity, and relativism, respectively, as classified in the Perry scheme.

Participants were asked to rate the degree of importance of each item on a four­point Likert scale. The results of the questionnaire were than transformed into LEP scores ranging from 200 to 500 (Moore, 1989), indicating Perry position two to five. The LEP questionnaire in Chi­nese has been developed through several revisions and pilot tests, and the internal consistency was found to be over 0.7 across the Perry posi­tions (Yang, 2005).

Although the LEP questionnaire is a convenient tool for large­scale samples (Moore, 1989), the raw LEP scores were only moderately corre­lated with scores obtained from a well­recognized written method – the measure of intellectual development (MID) that assessed Perry posi­tions by analyzing subjects’ written essays about educational experiences (Knefelkamp et al., 1984; Moore, 1989). Nevertheless, since the Perry positions obtained from LEP analysis and the MID written method were found to be significantly associated with each other (Moore, 1989), the constructs identified by the two instruments were conceptually related. Hence, the epistemological positions transferred from LEP scores were used as an indicator of epistemological development.

Results. The results of the study showed that the majority of the par­ticipants (over 70 percent) have not developed a complex form of episte­mology. The mean LEP scores (343.1, SD = 43.2) suggested that these students were at the stage of early multiplicity (position three in the Perry scheme). By definition, those who were in early multiplicity believe that most knowledge is known, and there exists a right way to find the correct answers. However, they start to realize that some knowledge domains are “fuzzy.” Early multiplicity actually signals the transition from a dualist to a mature multiplicity, in which individuals still carry

An epistemic framework for scientific reasoning134

some absolutist belief (Perry, 1970). The one­way ANOVA showed that the higher the LEP scores, the more the respondents showed in­depth understanding about the nature of evidence, and the more able they were to recognize that expert opinions need to be justified.

Content analysis of written responses regarding the evaluation of controversial information showed that while many students in the study relied heavily on concrete and numerical information for making judgments, the higher the Perry position, the more capable they were of identifying relevant evidence in making judgments. For example, when asked whether numerical data are the only type of evidence to draw con­clusions, seventeen students agreed while forty­six disagreed. Among those who disagreed, eight subjects (out of thirty­six) in position three stated that “information from various sources can also be used to draw conclusions.” However, only one (out of eight) in position two made such a comment. On the other hand, among those who agreed, three (out of seven) in position two mentioned that the numerical data are scientific, but no subjects in positions three or four made such a com­ment. Moreover, when students were asked if the issue should be settled by the numerical data alone, a greater proportion of students in position three expressed other concerns such as the safety of human lives and properties and the survival of the construction companies.

In this part of the study, only a few responses of students in posi­tion four were found. While these students stressed that science is a theory­based process of testing, and emphasized the use of direct and subjective data to support assertions, it was found that their reliance on experts was based on whether expert opinions were consistent with prior research or if experts provided reliable evidence. In other words, these students have started to think like evaluativists as defined by Kuhn (1991).

Study 4: The mode of informal reasoning and the personal epistemology of the sixth-graders (Yang et al., 2005)

Purpose. As mentioned, although the association between scientific rea­soning and personal epistemological beliefs has been stated in many studies (e.g., Driver et al., 2000; Kuhn, 1991; Simmons and Zeidker, 2003; Yang, 2004, 2005), whether the association fits in a developmen­tal trend could not be revealed unless more studies with young children were undertaken. Based on the studies with tenth­ and twelfth­graders as described previously, the main purpose of this study was to examine the relation between scientific reasoning in informal contexts and epis­temological perspectives exhibited by sixth­graders.

Fang­Ying Yang and Chin­Chung Tsai 135

Participants. Participants were forty­one children at the age of eleven and twelve with twenty­five females and sixteen males from eight classes in a suburban elementary school who were recommended by their teachers to participate in the interviews. To facilitate interview data­gathering about typical reasoning modes of sixth­graders, an average language ability and above­average academic performance were the main criteria for selecting subjects.

Measure. Similar to the design of previous studies, two science­ related issues with conflicting information were employed in the study. One issue concerned the feasibility of earthquake prediction and the other involved a protest event related to well­digging and land subsidence, which is the same issue examined in Study 2. The basic story of earth­quake prediction and interview questions are presented in Appendix C. During interviews, participants were asked to make comments on the certainty of expert knowledge and evaluate conflicting informa­tion presented in the two issues. Subjects were also required to reflect on their own ideas. Participants’ oral responses were analyzed based on concepts about the nature of knowledge and knowing defined by Kuhn’s model of epistemological understanding (Kuhn, 1999), Perry’s scheme (Perry, 1970), and the reflective judgment model proposed by King and Kitchener (1994).

Results. The major findings of the study are described as follows. First, when the participating children were asked whether, in any science research, agreement exists among experts, over 60 percent of the subjects expressed the absolutist view that experts uniformly agree about knowl­edge. About 30 percent of subjects were classified as using the multiplist form of epistemology due to the fact that they were able to recognize fuzziness in some knowledge domains. In addition, about 8 percent of the subjects who mentioned that agreement could be reached through discussions were classified as having the evaluativist epistemology.

Second, when children were asked to make a choice between two expert claims and provide reasons for their selections in the earthquake issue, most children (62 percent) recognized expert claims as a source to support or confirm personal beliefs. For others (38 percent), they speci­fied reasons or evidence provided by experts to justify personal ideas. The former type of view about expert claims matched with the pre­reflective view toward the process of knowing as defined in the reflec­tive judgment model (King and Kitchener, 1994), and also aligned with the absolutist view proposed by Kuhn (1999), whereas the latter type of view was frequently found in individuals with the multiplist form of epistemology (Kuhn, 1999) or the early quasi­reflective view as stated in the reflective judgment model (King and Kitchener, 1994).

An epistemic framework for scientific reasoning136

When discussing the issue of land subsidence, 68 percent of subjects held an absolutist idea that experts are the source of knowledge (King and Kitchener, 1994; Perry, 1970), or expert claims were assertions that could be correct or incorrect in their representation of reality (Kuhn, 1999). The rest of the children (32 percent) expressed a concept that claims could be justified by the authority figure. Such a concept reflects the early quasi­reflective judgment (King and Kitchener, 1994) which is parallel to the multiplicity stage in Perry’s scheme (Hofer and Pintrich, 1997). In short, our analyses indicated that an absolutist perspective toward the process of knowing was dominant among these sixth­graders.

Third, when evaluating conflicting information, it was found that about 41 percent of the students were able to identify evidence to support theory regarding the issue of earthquake prediction, while in the issue of land subsidence the successful rate dropped to 34 percent. In addition, children’s oral responses showed that elementary students practiced reflective thinking mainly to assure or defend their own thoughts. Chi­square analyses suggested that individuals’ views about the certainty of knowledge and concepts regarding evidence and expert opinions were associated with the coordination of theory and evidence in the earth­quake issue. It was found that those who held multiplist views performed better coordination of theory and evidence than the absolutists. How­ever, no such association was found in the land subsidence issue which provided a more complex context for reasoning than the earthquake issue. Such a result indicated that “context” or “complexity” of a prob­lem or issue could affect the performance of scientific reasoning.

Fourth, to see the effect of context on the displays of personal epis­temology, comparisons on the understanding about certainty of knowl­edge, source of knowledge, and concept about justification across issues were conducted. It has been mentioned that most students held an abso­lutist view about the certainty of knowledge in the earthquake issue. Although the certainty of knowledge was not specifically posited as an interview question in the land subsidence issue, children’s responses for why an authority claim was accepted or rejected indicated that these children displayed an early stage of reflective judgment in which indi­viduals assume knowledge as absolutely certain (King and Kitchener, 1994). Moreover, Chi­square analysis and symmetric measures showed only an approximately moderate association ( Chi­square = 5.63, Cramer’s V = 0.31, p < 0.1) for the understanding about source of knowledge identified in the land subsidence issue and concepts about justification assessed in the earthquake issue.

If personal epistemology is context­independent, the display of per­sonal epistemology should remain consistent regardless of different

Fang­Ying Yang and Chin­Chung Tsai 137

issues. Consequently, a significant and strong association between beliefs about the process of knowing should be expected across issues. Hence, the approximate association found in this study indicated that context may have some impact on the display of personal epistemology. In addition, content analysis indicated that different contexts seemed to activate different justification criteria for making judgments. Given that beliefs about the process of knowing were only approximately asso­ciated with each other across issues, and different justification criteria were employed in different issues, it was concluded that context could be a significant factor affecting the display of personal epistemology, especially in the dimension of justification for knowing. It should also be noted that an additional study has found similar results (Yang and Tsai, in press).

Study 5: The mode of informal reasoning and the personal epistemology of the eighth-graders

Purpose and design. Although the previously mentioned studies have seemingly supported the developmental link between scientific reason­ing and personal epistemology, more evidence from students of various ages is needed to build a comprehensive framework. Hence, a recent study with middle­school subjects was conducted and is reported here. This study employed a similar research method and procedure to that described in Study 4. Instead of interviews, an open­ended question­naire which contained all the interview material and content was used to assess epistemological understanding and scientific reasoning. Par­ticipants were seventy­four eighth­grade learners. Some coding exam­ples of written responses are presented in Appendix D.

Results. Analysis of students’ responses to questions regarding agree­ment among experts showed that about 34 percent of eighth­graders were identified as absolutists or dualists, while 51 percent were multiplists regarding beliefs about certainty of knowledge. A few students (15 per­cent) were found to have evaluativist views that agreement can be reached through discussions. As far as belief about source of knowledge was con­cerned, 52 percent of participants thought that expert claims could sup­port or defend personal ideas, while the rest (48 percent) recognized that personal ideas could be justified by expert opinions when reasoning about the earthquake issue. However, in the land subsidence issue, over 75 per­cent of subjects turned to the absolutist concept that expert claims were assertions that could be wrong, or experts provided certain knowledge.

As far as criteria for judging information were concerned, content analysis found that the frequently mentioned concerns during responses

An epistemic framework for scientific reasoning138

regarding the earthquake issue were: the existence of supporting evi­dence, the coherence between information and personal beliefs or theo­ries, the relevance of the information to the problem to be solved, the reliability of the research equipment, and the repetition of evidence (for example, some respondents mentioned that “the astronomy group should conduct more observations to see if the result is the same …”). When presented with the land subsidence issue, the eighth­graders considered mostly the relevance of the information to the problem to be solved and the reliability of the information source. Seemingly, the socio­scientific issue is multidimensional in the sense that the decision­making on this kind of issue also required considerations of human lives and social aspects of science would induce different epistemologi­cal standpoints.

As for the coordination of theory and evidence, only 24 percent of the subjects identified evidence to support theory in discussing the earthquake­prediction issues, while 27 percent did so in the land sub­sidence issue. The similar percentage distributions in responses to the two issues seemed to suggest that problem context did not affect the coordination of theory and evidence. On the topic of reflective reason­ing, 70 percent of participants showed high confidence toward their own ideas which, as mentioned in Study 4, implied absolutist epistemology. Contrary to Study 4, Chi­square analysis did not find any association between personal epistemology and scientific reasoning in either the earthquake or land subsidence issue.

Summaries of previous studies

For the studies with high school subjects, Study 1 proposed that sci­entific reasoning in informal contexts could be mediated by personal epistemological beliefs. In Studies 2 and 3, multiplicity, particularly the early form, was found to be the dominant form of personal epistemol­ogy among high school tenth­grade subjects. It was also found in Study 2 that performance of scientific reasoning regarding coordination of theory and evidence and reflective reasoning in informal contexts could be explained by personal epistemology. In addition, the result of Study 3 supports that the personal epistemological understanding in science correlates with the developmental levels of personal epistemol­ogy. Accordingly, it was concluded that associations between scientific reasoning and personal epistemology were apparent among the high school learners.

Study 4 revealed that sixth­graders were mostly absolutists and dis­played lower ability than the high school subjects in performing scientific

Fang­Ying Yang and Chin­Chung Tsai 139

reasoning in informal contexts. Such a finding seemed to suggest a link between developmental tendency for scientific reasoning and personal epistemology. Nevertheless, findings of Study 5 with eighth­ graders did not align with the developmental trend as expected. It appeared that the displays of personal epistemology were unstable across the two issues, and the performance of coordination of theory and evidence in discussing issues was weakened among the eighth­graders in com­parison with that of sixth­graders. In addition, context did not seem to affect the practice of scientific reasoning. Moreover, contrary to the studies with elementary and high school subjects, Chi­square analysis did not show any significant association between scientific reasoning and personal epistemology among eighth­graders.

The results of Study 5 seemed to differ from that of Mason and Scir­ica’s (2006) recent study. They assessed eighth­graders’ epistemologi­cal understandings by a domain­general instrument and showed that over 74 percent of students were characterized as multiplists while 26 percent were evaluativists. Assessed by interviews, the displays of per­sonal epistemology shown in Study 5 were found to vary across differ­ent issues. In Mason and Scirica’s (2006) study, analysis of the relation between epistemological understanding and use of argumentation skills (including formulating arguments, counterarguments, and rebuttal in reasoning about two pure­science controversial issues) showed that epistemological understanding was a predictor for the performance of argumentation skills. However, Study 5 found no association between epistemological beliefs and scientific reasoning. Although results of the two studies seemed to contradict each other, it should be noted that the aspects of personal epistemology and reasoning skills investigated in the two studies were not exactly the same. While Mason and Scirica (2006) examined personal epistemology with a context­free question­naire, our study assessed students’ epistemological perspectives in dif­ferent contexts. In addition, instead of investigating all argumentation skills as mentioned above, our study focused on reasoning compe­tences such as the coordination of theory and evidence and reflective reasoning. Therefore, more studies are needed to clarify the apparent contradictions.

Our previous studies suggest that the development of scientific rea­soning and the development of personal epistemological beliefs might not be a linearly progressive process. To further discuss the perform­ance of scientific reasoning and personal epistemological perspectives across different age groups, an explicit comparison among research findings of the previously mentioned studies is presented in the following section.

An epistemic framework for scientific reasoning140

Cross-comparisons on the performances of scientific reasoning in informal contexts and personal epistemology across various age groups

Beliefs about the nature of knowledge across different age groups

For studies regarding belief about the nature of knowledge, investi­gations with sixth­ and eighth­grade learners (Studies 4 and 5) were conducted in the context of an earthquake issue. Since the issue of earthquake prediction is controversial and not discussed in the school curricula, it was believed that students should supposedly activate gen­eral epistemic beliefs in reasoning (Muis et al., 2006; Songer, 1989). Moreover, personal epistemology discussed in Study 3 was also con­sidered as a domain­general construct, which was assessed by the LEP questionnaire. Hence, the comparison of personal epistemology con­cerning the nature of knowledge across three groups of students were made as shown in Table 5.2, for the purpose of showing the general tendency.

As Table 5.2 shows, most elementary students held an absolutist form of epistemology while the tenth­graders were mostly identified as the multiplists in accordance with the LEP scores. Examining the fre­quency distributions across age, it was found that the number of abso­lutists decreased and multiplists increased with the increase of age. The percentage distributions suggested a progressive development of the epistemological perspective concerning the nature of knowledge from absolutist to multiplist. Nevertheless, while the percentage of evaluativ­ist increased at the junior level (15 percent), it fell back at the senior level (9 percent). Such a non­linear trend has previously been discussed in many epistemological studies. For instance, Perry (1970) mentioned that the development of personal epistemology is a dynamic process, which may move back and forth among adjacent positions if the thinker has not reached a mature stand.

By reviewing several longitudinal studies, Chandler et al. (2002) con­cluded that epistemological development could be recursive, and there­fore regression to earlier stages is possible. Considering the changes of personal epistemology in different knowledge domains and educational contexts, Bendixen and Rule (2004) proposed a spiral­like rather than linear developmental model for personal epistemology. Given that the majority of senior students displayed the multiplist view, which is seem­ingly the stable form of epistemology in this age, the relatively higher distributions of evaluativists and absolutists found in junior learners imply that the eighth­graders were likely entering a rather unstable stage of epistemological development.

Fang­Ying Yang and Chin­Chung Tsai 141

Table 5.2. Percentage distributions of epistemological perspectives regarding the nature of knowledge (certainty of knowledge) among students of different grade levels

Grade

Belief about certainty of knowledge

Absolutist Multiplist Evaluativist

The sixth­graders (elementary learners) 63% 29% 8%

The eighth­graders (junior learners) 34% 51% 15%

The tenth­graders (senior learners) 20% 71% 9%

Beliefs about the processes of knowing across different age groups

The investigation of beliefs about the process of knowing was focused on revealing learners’ views toward evidence and expert opinion. Since Study 3 with high school subjects considered a different issue, the com­parison was made only between the sixth­ and eighth­grade groups. As suggested in Table 5.3, a developmental trend seemed to appear more strongly in reasoning on the science­based issue (the earthquake predic­tion). While the percentages of students’ epistemological perspectives about the nature of knowing displayed by sixth­graders remained stable across issues, there was a fallback from the multiplist to the absolutist in the eighth­grade group when reasoning about the socio­scientific issue. For the middle­school learners, different issues seemed to activate dif­ferent epistemological perspectives in reasoning.

Table 5.3. Percentage distributions of epistemological perspectives regarding the process of knowing (source of knowledge) among students of different grade levels

Belief about evidence and expert information

Science­based issue Socio­scientific issue

Grade Absolutist Multiplist Absolutist Multiplist

The sixth­graders (elementary learners)

62% 38% 68% 32%

The eighth­graders (junior learners)

52% 48% 75% 25%

An epistemic framework for scientific reasoning142

The issue regarding change of personal epistemology in different contexts and/or domains has been discussed in many psychological studies. Some scholars argue that the change is a matter of developmen­tal phenomenon (Bendixen and Rule, 2004; Hofer, 2000; Muis et al., 2006), while others mentioned that problems with different epistemo­logical roots might activate different epistemological resources (Louca et al., 2004). In our studies, the effect of context was more apparent in our eighth­grade subjects. This finding supports our argument that the junior learners were unstable in their epistemological development. However, since the analysis did not include tenth­grade learners, the inference might be limited in its generalizability.

Coordination of theory and evidence and reflective thinking across different ages

Table 5.4 lists the performances of scientific reasoning regarding the coordination of theory and evidence by learners in different age groups. It should be noted that the coordination of theory and evidence per­formed by tenth­graders was assessed only for the socio­scientific issue about land subsidence (Study 2). As Table 5.4 shows, while perform­ances of coordination of theory and evidence displayed by learners at elementary and high school levels suggest an overall progressive trend, there seemed to be a drawback at the junior high level. While Chi­square and ANOVA analyses found statistical associations between scientific reasoning and personal epistemology among sixth­ and tenth­graders, no such relation was found among the junior subjects.

Table 5.4. Performances of coordinating theory and evidence across different issues

Identifying the role of evidence to support a theory

Science­based issue Socio­scientific­based issue

Grade Success Fail Success Fail

The sixth­graders (elementary learners)

41% 59% 34% 66%

The eighth­graders (junior learners)

24% 76% 27% 73%

The tenth­graders (senior learners)

N.A. N.A. 57% 43%

Fang­Ying Yang and Chin­Chung Tsai 143

The result suggested that ability to coordinate theory and evidence by junior learners did not go along with the development of personal epistemology.

As far as reflective thinking is concerned, which was assessed in the land subsidence issue, Table 5.5 shows that the majority of both sixth­ and eighth­graders as presented in Studies 4 and 5 exhibited high confidence toward their own ideas, while most of the senior learn­ers expressed uncertainty toward personal thoughts. Kuhn (1991) mentioned that absolutists tended to show high confidence in the cer­tainty of their personal beliefs. Accordingly, the finding confirmed that elementary and junior school learners held a more primitive form of epistemological belief. Moreover, while no association was found between belief about the certainty of knowledge and reflective reason­ing, Chi­square analysis, as shown in Study 4, found that children in the sixth­grade who held absolutist views toward the process of know-ing tended to be more certain toward their thoughts than those who were identified as multiplists. However, no similar association was found among eighth­graders in Study 5. Although the data for the senior learners were not tested by the Chi­square method in this part of the study, a similar association could be expected because correla­tion between coordination of theory and evidence and views toward the process of knowing was found among high school students in Study 3, and the performances of coordination of theory and evidence and reflective reasoning was found to be consistent in Study 2. Accord­ingly, it is concluded that junior learners were likely to be unstable in the development of personal epistemology and scientific reasoning. In addition, the previous findings also suggest that the performance of

Table 5.5. Performance of reflective thinking in discussing a socio-scientific issue (land subsidence issue)

Reflective thinking

Grade Sure of personal idea Unsure of personal idea

The sixth­graders (elementary learners)

70% 30%

The eighth­graders (junior learners)

70% 30%

The tenth­graders (senior learners)

14% 86%

An epistemic framework for scientific reasoning144

reflective thinking could be more related to students’ epistemological belief about the process of knowing.

Effects of personal epistemology on scientific reasoning

The overall effects of personal epistemology on scientific reasoning are summarized in Table 5.6. The comparisons made in Table 5.6 seemed to point to an inconsistency with the general developmental model. Such a result can be partially explained by Perry’s developmental scheme for personal epistemology. As mentioned before, Perry (1970) argued that the development of personal epistemology may move back and forth among adjacent positions if the thinker has not reached a mature stand. This could be the case for the junior subjects (the eighth­graders) given that the percentage distributions of personal episte­mology concerning nature of knowledge hardly pointed to a strongly dominating one, compared to those demonstrated by the elementary and senior subjects. That is, although many of the junior learners were

Table 5.6. Effects of personal epistemology on scientific reasoning regarding coordination of theory and evidence and reflective reasoning

Grade

(a) Effect of personal epistemology on coordination of theory and evidence

The effect across contexts

The fifth­ and sixth­graders (elementary learners)

Found Consistent

The eighth­graders (junior learners)

Not found No effect found

The tenth­graders (senior learners)

Found N.A.

Grade

(b) Effect of personal epistemology on reflective thinking

The effect across contexts

The fifth­ and sixth­graders (elementary learners)

Found Consistent

The eighth­graders (junior learners)

Not found No effect found

The tenth­graders (senior learners)

N.A. N.A.

Fang­Ying Yang and Chin­Chung Tsai 145

multiplists (51 percent), still 34 percent of students were categorized as absolutists and 15 percent evaluativists, as shown in Table 5.2. In addition, a regression of belief about the process of knowing was found among the eighth­graders when they were placed in the land subsidence issue. It is thus conjectured that, affected by the unstable development of epistemological beliefs, junior learners’ performance of scientific reasoning in informal contexts became somewhat unpredictable. In addition to the influence of personal epistemology, context may also add to the complexity of reasoning. As the percentage distributions in Table 5.4 show, the difficulty of scientific reasoning was more apparent in the land subsidence issue, which is multidimensional by nature, than in the issue of earthquake prediction, which is more related to domain­specific knowledge. In conclusion, our studies suggested that personal epistemology and context interact with scientific reasoning in informal contexts.

An epistemic framework for the development of scientific reasoning

Some psychologists have proposed theoretical frameworks regarding the dynamic nature and developmental variables of personal episte­mology (e.g., Bendixen and Rule, 2004; Muis et al., 2006; Schommer­Aikins, 2004). Consistent with these frameworks, our framework aims to describe the dynamic associations between the performance of sci­entific reasoning in informal contexts and the development of personal epistemology. As illustrated in Figure 5.1, our framework emphasizes the idea that development of personal epistemology is a changing proc­ess which results in variations in the performance of scientific reason­ing in informal contexts. In turn, reasoning experiences will contribute to the development of personal epistemology. Moreover, the reasoning performance is influenced by the complexity of context of the encoun­tered issue.

As discussed in the previous section, personal epistemology was found to develop with age and educational experiences. However, the process of progression was not linear. As demonstrated in Figure 5.1, we propose that the development of personal epistemology might undergo three kinds of processes including reinforcement, withdrawal, and advancement of epistemological position. Given the associations between personal epistemology and scientific reasoning in informal contexts found in previous studies, it was suggested that the develop­ment of personal epistemology would give rise to three consequences of scientific reasoning in informal contexts, namely stable performance,

An epistemic framework for scientific reasoning146

withdrawal, and improvement of reasoning. The proposed relations are illustrated as follows.

First, when an individual reaches a stable epistemological position, his/her performance of scientific reasoning in informal contexts will stably correspond to his/her epistemological positions. For example, an absolutist will rely on expert opinions as evidence to support their favored theory, while a multiplist will consider whether the data pro­vided by experts actually agree with their personal theory. Evaluativist thinkers would examine whether a claim or theory is supported by evi­dence that is widely accepted by experts in the relevant area. The stable performance of scientific reasoning will reinforce the existing personal epistemology.

Second, our studies on elementary, middle, and high school students showed that during development, there seemed to be a period of time when personal epistemology withdraws to a previous stage and the

Stable performance of scientific reasoning

Stableor not

Personal epistemologyin developing

Low

Complexityof context4

High

Withdrawal in scientific reasoning

Improvement inscientific reasoning

YES

NO

Reinforcement 1

Advancement 3

Withdrawal 2

1. Reinforcement process results in the maintenance of current epistemological position.2. Withdrawal process results in the fallback of epistemological position. 3. Advancementprocess results in the advance of epistemological position. 4. The complexity of contextaffects the reasoning performance.

Figure 5.1: Theoretical framework for the development of scientific reasoning in informal contexts and personal epistemology

Fang­Ying Yang and Chin­Chung Tsai 147

performance of scientific reasoning becomes poorer and unpredictable. The fallback of personal epistemology and scientific reasoning, as the eighth­graders displayed in Study 5, could be a result of developmental transition in which a thinker’s personal epistemology has just developed to an unstable stage which consequently brings about difficulties in scientific reasoning. The negative experiences in reasoning then cause a withdrawal of development in personal epistemology.

Third, on some occasions (for example, reasoning on a difficult issue with help from teachers), the thinker might perform scientific reasoning in a way that is advanced relative to his/her current episte­mological status. The improvement of scientific reasoning provides an opportunity for the thinker to reflect on their own beliefs about knowledge and knowing, which will then promote further develop­ment of personal epistemology to the next stage. Such an improve­ment process for scientific reasoning and personal epistemology is well supported by Vygotsky’s idea of “zone of proximal development” (Vygotsky, 1978).

In addition to the reciprocal interactions between personal epis­temology and scientific reasoning, the complexity of context of the encountered issue would also intervene during the process of develop­ment. For example, as shown in Studies 4 and 5, when the issue to be solved is multidimensional in nature, all opinions might seem rational to multiplists. As a result, they would have difficulty applying diverse epistemological criteria in reasoning about the complex issue (context of high­complexity). Consequently, the thinker would withdraw to an earlier epistemological standpoint (i.e., absolutist view) which encour­ages him/her to find a right answer from experts. On the other hand, when the complexity of context of the encountered issue is low, the performance of scientific reasoning remains steady.

In short, many researchers (e.g., Kuhn, 1991, 1999; Tsai, 2004) have proposed that the practice of argumentative reasoning provides opportunities for thinkers to reflect on their epistemological beliefs; our dynamic framework hypothesizes that reasoning practices in vari­ous problem contexts may in turn contribute to the development of personal epistemology.

Research implications

According to our framework, further research is necessary to explore the performances of scientific reasoning in various problem contexts by students in different epistemological positions. For example, stud­ies across different ages or follow­up longitudinal investigations and

An epistemic framework for scientific reasoning148

experimental designs such as reasoning tasks, with or without help from teachers, are channels to observe students’ improvement of scientific reasoning and the development of personal epistemology. In addition, in our studies only a few subjects were characterized as evaluativists. As a result, we lacked sufficient data to depict the reasoning behav­ior of evaluativists and to draw relations between evaluativists’ reason­ing performance and personal epistemology. Hence, there is a strong need to inspect subjects in the advanced epistemological position. It is thus suggested that similar studies as described in this chapter can be extended to graduate students, who theoretically should have developed more evaluativist views.

Moreover, as presented in our analysis and also mentioned by many researchers, the development of personal epistemology while progress­ing overall with age may undergo puzzling periods in which personal epistemology jumps back and forth before settling down in the next stage (e.g., Bendixen and Rule, 2004; Chandler et al., 2002; Muis et al., 2006; Perry, 1970). In our studies, this puzzling period occurred at the middle­school level. The junior learners were described as being unstable because a strong dominating epistemological standpoint about the nature of knowledge could not be detected at this level in contrast to the sixth­ and tenth­grade levels, and a regression of belief about the process of knowing was observed when the problem context became complicated. In addition, junior subjects displayed poor performance of scientific reasoning in informal contexts. Hence, to have a better understanding about the non­linear progression of epistemological development, more studies with junior subjects are needed.

Another issue worth further exploration is the culture effect. Muis and colleagues (2006) proposed that when students enter middle or high school, the changes of educational content and environments could trigger a regression of epistemological development. In Mason and Scirica’s study (2006), the performance of argumentation skills by eighth­graders could be predicted by personal epistemological under­standing. However, it was not the case in our study. Although our eighth­grade subjects were not freshmen in middle school, the changes of teaching and learning environment in grade eight is actually more dramatic than the beginning grade (grade seven). In the education sys­tem of Taiwan, students in the eighth­grade need to start the prepa­ration for the entrance examination for high schools. As a result, the teacher­centered instruction which aims to transmit factual knowledge dominates the eighth­grade classrooms. Consequently, students would become more familiar with the absolutist perspective about the process of knowing. In contrast, various instructional approaches can be seen

Fang­Ying Yang and Chin­Chung Tsai 149

in the sixth­ and tenth­grade classrooms because of less pressure from the national standardized examination. Such a situation may be applied to many Asian countries. Accordingly, the issue of cultural impact on the development of personal epistemology that has recently started to gain attention (e.g., Haerle and Bendixen, 2008; Hofer, 2005; Khine, 2008; Muis et al., 2006; Qian and Pan, 2002; Schommer­Aikins, 2004) should be extensively studied. As such, similar studies should be con­ducted with subjects of different cultural and social backgrounds.

Educational implications

In this chapter, we cross­analyzed the performances of scientific rea­soning in informal contexts and the development of personal epistemol­ogy among sixth­, eighth­, and tenth­graders. An epistemic framework of scientific reasoning is then proposed to illustrate the reciprocal asso­ciation between scientific reasoning in informal contexts and personal epistemology. As shown in our framework (see Figure 5.1), while per­sonal epistemology mediates the performance of scientific reasoning, reasoning experiences in turn will help (or hinder) the advancement of personal epistemology. In science education, many educators have advocated for the importance of argumentation activities in science classrooms to promote the development of scientific reasoning (Driver et al., 2000; Duschl and Osborne, 2002). Nevertheless, it has not been widely applied to classroom practice. Linking the need for argumen­tation activities in science classrooms to studies in the development of personal epistemology, our framework suggests that the practice of scientific reasoning in various contexts will not only encourage the mastery of argument skills, but also activate different epistemologi­cal beliefs. This is consistent with the suggestion made by others (e.g., Kuhn, 1991, 1999; Tsai, 2004) that it is helpful to provide opportuni­ties for examination and reconciliation of different epistemic beliefs. Hence, according to the advancement process shown in Figure 5.1, it is recommended that to promote the development of scientific reason­ing and further advanced personal epistemology, instructors of science (and also of many other academic domains) need to expose learners to argumentation activities in informal contexts. Notably, for learn­ers who are more in a transition of epistemological development, such as the eighth­graders discussed in above study, instructors need to be more cautious in using complex issues for classroom activities to avoid a withdrawal in reasoning and personal epistemology.

According to our studies and the proposed theoretical framework, practicing scientific reasoning in informal contexts can be achieved by

An epistemic framework for scientific reasoning150

exposing children in groups to anomalous data or conflicting informa­tion in science (or other academic domains), which allow students to socially construct scientific knowledge, and to evaluate critically the relation between theory and evidence. As discussed in our epistemic framework, we theorized that improvement in scientific reasoning, which can be achieved by the help of teachers, will in turn advance personal epistemology (i.e., advancement in our framework). Hence, it is important that teachers provide proper scaffolding to guide stu­dent discourses (Duschl and Osborne, 2002). Gradually, students can master argument skills and understand the epistemology underlying the argument skills (Hammer and Elby, 2002). Furthermore, teachers are recommended to create situations or opportunities for individual students to explicitly examine their own views about what knowledge is and how knowledge is constructed.

Based on our theoretical framework, some detailed instructional recommendations regarding argumentation activities in science class­rooms are described in the following sections. For sixth­graders who are mostly absolutists, the instructional emphasis can be placed on introducing multiplist epistemology by discussing the uncertainty that exists in science (or some specific discipline). Pure­science in­dispute issues such as earthquake prediction and dinosaur extinction are suit­able material to show that different expert groups might not agree with each other and that each group has found evidence to support their claims. At this point, it is no surprise that these children, being absolut­ists, might be eager to know the winner group. Instead of simply tell­ing them the widely accepted theory, teachers should help students in groups to examine the merits of evidence found in the various groups so that young learners can acquire argument skills step­by­step and eventually gain deeper epistemological understanding about theory and evidence (i.e., advancement in our framework). In addition, since the complexity of context may affect the display of personal epistemol­ogy and the performance of scientific reasoning in informal contexts, it is suggested that teachers should carefully move the problem context from low to high levels of complexity to avoid failure in reasoning tasks that might hinder the development of personal epistemology (i.e., with-drawal in our framework).

As our studies showed, most high school students have developed a multiplist epistemology. Thus, to enhance the practice of making argu­ments based on accountable evidence it is important for them to under­stand the relation between theory and evidence. To prepare for further development of the evaluativist epistemology, classroom activities should particularly emphasize the evaluations of theories (arguments)

Fang­Ying Yang and Chin­Chung Tsai 151

based on the accountability of evidence. In addition, for this age level, we think it is more appropriate to discuss in­depth the social process of knowledge construction. These activities aim to promote sophisticated scientific reasoning and further stimulate the progression of personal epistemology (i.e., advancement in our framework).

Science teachers can start instruction by discussing the history of science that concerns the development of currently accepted public knowledge, such as plate tectonics in earth science or the theory of evolution in biology. These theories were not accepted when they were first introduced to the scientific community. As advocated by some science educators (Duschl, 1990; Matthews, 1992), learners can learn from history how a theory is constructed, judged, and eventu­ally accepted in the scientific community. Classroom activities may also include critical discussion of the criteria, such as logical, empiri­cal, sociological, and historical criteria (Root­Bernstein, 1984), used in scientific communities for judging evidence and provide evidence of the role of the scientific community in reaching consensus judgments. After the discussion on accepted public knowledge, instructors can move to pure­science in­dispute issues, and ask students to use the judgmental criteria they learned previously to analyze the conflicting theories. Furthermore, to acquire the evaluativist understanding that knowledge is contextual, we recommended that reasoning and evalu­ation tasks for high school students should be carried out in a variety of problem contexts. This is proposed based on the idea that complex­ity of context mediates reasoning performance and further stimulates the advancement of personal epistemology (i.e., advancement in our framework). Thus, classroom discussions and/or discourses can be moved on to socio­scientific issues that allow learners to understand that theories or claims and judgmental criteria could change depend­ing on the nature of the problems.

Moreover, adolescent learners need to be given opportunities to reflect on their own beliefs about knowledge and knowing, and com­pare their personal beliefs to the construction of public knowledge. Reflections on one’s own beliefs are critical for the development of an evaluative epistemology because, as Kuhn (1999) pointed out, the core dimension driving progression of epistemological understanding is the coordination of the subjective and objective components of knowing. As discussed in our studies, reflective reasoning is also an important aspect of scientific reasoning. The improvement of reflective reason­ing is thus expected to promote further development of epistemological understanding (i.e., advancement in our framework). Activities such as writing journals, peer sharing of personal beliefs, and peer discussions

An epistemic framework for scientific reasoning152

about the development of public knowledge are useful ways to promote self­reflection.

As discussed before, the eighth­graders who are likely to be in the transition between absolutists and multiplists displayed unstable epis­temological positions and reasoning performance. To promote the development of personal epistemology and scientific reasoning, it is suggested that the practice of argument skills might be initially placed in formal contexts (that is, a low complexity of context) in which no controversy would be raised to puzzle these learners (i.e., taking care­ful consideration of context complexity to avoid withdrawal mentioned in our framework). Gradually, teachers can start to situate learners in more complex (but not too complex) issues such as earthquake prediction and dinosaur extinction as described for the elementary students. It should be noticed that even though the eighth­graders may exhibit multiplist epistemology after argumentation activities, teachers should keep in mind that these students might fall back to an absolutist stance when the issues they encounter become compli­cated. Therefore, middle school students need to be engaged in more argumentative activities to secure their epistemological position and reasoning skills (i.e., reinforcement in our framework). In addition, to promote the development of personal epistemology, students at this stage need to be encouraged to reflect on their beliefs about knowl­edge and knowing. Instead of comparing personal beliefs to the con­struction of public knowledge as is recommended for the high school students, the reflection should be focused on the comparison of per­sonal epistemological beliefs positioned in different problem contexts. In this way, students can recognize the variations of their own beliefs. Thus, the reconciliation among personal beliefs becomes possible. As pointed out by our theoretical framework, the self­reflection on personal beliefs is expected to trigger the advancement of personal epistemology. Writing journals, teacher probing, and peer discussions regarding personal beliefs in different problem contexts are helpful to promote such self­reflection.

As mentioned before, scientific reasoning habits are important in daily experiences because they provide important ways to make rational and sound judgments about controversial issues in social contexts. One of the challenges presented to science educators and perhaps to all edu­cators is how to create learning situations where these powerful forms of thought can be anchored in situations that may be generalizable beyond science classrooms. In this chapter, we proposed a theoretical framework that points out the close links between scientific reason­ing and personal epistemology across different age groups. Based on

Fang­Ying Yang and Chin­Chung Tsai 153

the framework, we present several practical instructional recommen­dations to enhance science classroom learning. It is hoped that these recommendations can inspire further instructional considerations in promoting scientific reasoning habits and the development of students’ personal epistemology.

R EF ER ENCES

Anderson, O. R. and Demetrius, O. J. (1993). A flow­may method of repre­senting cognitive structure based on respondent’s narrative using science content. Journal of Research in Science Teaching, 30, 935–69.

Astington, J. W., Pelletier, J., and Homer, B. (2002). Theory of mind and epis­temological development: The relation between children’s second­order false­belief understanding and their ability to reason about evidence. New Ideas in Psychology, 20, 131–44.

Baron, J. (1988). Thinking and deciding. Cambridge: Cambridge University Press.

Baxer Magolda, M. B. (1992). Knowing and reasoning in college: Gender- related patterns in students’ intellectual development. San Francisco: Jossey­Bass.

Bendixen, L. D. and Rule, D. C. (2004). An integrative approach to personal epistemology: A guiding model. Educational Psychologist, 39, 31–42.

Burr, J. E. and Hofer, B. K. (2002). Personal epistemology and theory of mind: Deciphering young children’s beliefs about knowledge and know­ing. New Ideas in Psychology, 20, 199–224.

Chandler, M. J., Hallett, D., and Sokol, B. W. (2002). Competing claims about competing knowledge claims. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (145–68). Mahwah, NJ: Lawrence Erlbaum Associates.

Driver, R., Newton, P., and Osborne, J. (2000). Establishing the norms of sci­entific argumentation in classrooms. Science Education, 84, 287–312.

Duschl, R. (1990). Restructuring science education. New York: Teachers College Press.

Duschl, R. A. and Osborne, J. (2002). Supporting and promoting argumen­tation discourse in science education. Studies in Science Education, 38, 39–72.

Giere, R. N. (1991). Understanding scientific reasoning. Forth Worth, TX: Holt, Rinehart and Winston.

Hammer, D. and Elby, A. (2002). On the form of a personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (169–90). Mahwah, NJ: Lawrence Erlbaum Associates.

Haerle, F. C. and Bendixen, L. D. (2008). Personal epistemology in elementary classrooms: A conceptual comparison of Germany and the United States and a guide for future cross­cultural research. In M. S. Khine (Ed.), Knowing, knowledge and beliefs: Epistemological studies across diverse cultures (151–76). Netherlands: Springer.

An epistemic framework for scientific reasoning154

Hofer, B. K. (2000). Dimensionality and disciplinary differences in personal epistemology. Contemporary Educational Psychology, 25, 378–405.

(2001). Personal epistemology research: Implications for learning and instruction. Educational Psychology Review, 12, 353–82.

(2005). Developing culturally inclusive models of epistemic beliefs: A roundtable discussion. Paper presented at the meeting of the American Educational Research Association, Montreal, Canada, April.

Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learn­ing. Review of Educational Research, 67, 88–140.

Hofer, B. K. and Pintrich, P. R. (Eds.) (2002). Personal epistemology: The psy-chology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erl­baum Associates.

Jiménez­Aleixandre, M. and Pereiro­Munoz, C. (2002). Knowledge produc­ers or knowledge consumers? Argumentation and decision making about environmental management. International Journal of Science Education, 24, 1171–91.

Kelly, G. J. and Duschl, R. A. (2002). Toward a research agenda for epistemologi-cal studies in science education. Paper presented at the Annual Meeting of the National Association of Research in Science Teaching, New Orleans, LA, April.

Khine, M. S. (Ed.) (2008). Knowing, knowledge and beliefs: Epistemological stud-ies across diverse cultures. Netherlands: Springer.

King, P. and Kitchener, K. (1994). Developing reflective judgment: Understand-ing and promoting intellectual growth and critical thinking in adolescents and adults. San Francisco: Jossey­Bass.

Kitchener, K. S. (1983). Cognition, metacognition and epistemic cognition. Human Development, 26, 222–32.

Knefelkamp, L., Fitch, M., Taylor, K., and Moore, W. (1984). The measure of intellectual development: A review and update. Paper presented at the annual meeting of the American College Personnel Association, Baltimore, MD, April.

Kortland, K. (1996). An STS case study about students’ decision making on the waste issue. Science Education, 80, 673–89.

Kuhn, D. (1991). The skill of argument. Cambridge: Cambridge University Press.

(1993). Science as argument: Implications for teaching and learning scien­tific thinking. Science Education, 77, 319–37.

(1999). A developmental model of critical thinking. Educational Researcher, 28, 16–46.

Kuhn, D. and Weinstock, M. (2002). What is epistemological thinking and why does it matter? In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing (121–44). Mahwah, NJ: Lawrence Erlbaum Associates.

Kuhn, T. S. (1970). The structure of scientific revolutions. Chicago: The Univer­sity of Chicago Press.

Fang­Ying Yang and Chin­Chung Tsai 155

Lawson, A. E., Clark, B., Cramer­Meldrum, E., Falconer, K. A., Kwon, Y. J., and Sequist, J. M. (2000). The development of reasoning skills in college biology: Do two levels of general hypothesis­testing skills exist? Journal of Research in Science Teaching, 37, 81–101.

Lee, K. and Homer, B. (1999). Children as folk psychologists: The develop­ing understanding of mind. In A. Slater and D. Muir (Eds.), The Black-well reader in developmental psychology (228–52). Malden, MA: Blackwell Publishers.

Louca, L., Elby, A., Hammer, D., and Kagey, T. (2004). Epistemological resources: Applying a new epistemological framework to science instruc­tion. Educational Psychologist, 39, 57–68.

Mansfield, A. F. and Clinchy, B. M. (2002). Toward the integration of objec­tivity and subjectivity: Epistemological development from 10 to 16. New Ideas in Psychology, 20, 225–62.

Mason, L. and Scirica, F. (2006). Prediction of students’ argumentation skills about controversial topics by epistemological understanding. Learning and Instruction, 16, 492–509.

Matthews, M. R. (1992). History, philosophy, and science teaching: The present rapprochement. Science and Education, 1, 11–47.

Means, M. L. and Voss, J. F. (1996). Who reasons well? Two studies of infor­mal reasoning of different grade, ability, and knowledge levels. Cognition and Instruction, 14, 139–78.

Moore, W. S. (1989). The “learning environment preferences”: Exploring the construct validity of an objective measure of the Perry scheme of intellec­tual development. Journal of College Student Development, 30, 504–14.

Muis, K. R., Bendixen, L. D., and Haerle, F. C. (2006). Domain­generality and domain­specificity in personal epistemological research: Philosophi­cal and empirical reflections in the development of a theoretical frame­work. Educational Psychology Review, 18, 3–54.

Newton, P., Driver, R., and Osborne, J. (1999). The place of argumentation in the pedagogy of school science. International Journal of Science Education, 21, 553–76.

Perkins, D. N. (1985). Reasoning as imagination. Interchange, 16, 14–26. (1989). Reasoning as it is and could be: An empirical perspective. In

D. M. Topping et al. (Eds.), Thinking across cultures: The third interna-tional conference on thinking (175–94). Hillsdale, NJ: Lawrence Erlbaum Associates.

(1991). Everyday reasoning and the roots of intelligence. In J. F. Voss et al. (Eds.), Informal reasoning and education (83–106). Hillsdale, NJ: Lawrence Erlbaum Associates.

Perkins, D. N. and Salman, G. (1989). Are cognitive skills context bound? Educational Researcher, 18, 16–25.

Perry, W. G. (1970). Forms of intellectual and ethical development in the college years. San Francisco, CA: Jossey­Bass.

Pintrich, P. R. (2002). Future challenges and directions for theory and research on personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.),

An epistemic framework for scientific reasoning156

Personal epistemology: The psychology of beliefs about knowledge and knowing (389–414). Mahwah, NJ: Lawrence Erlbaum Associates.

Qian, G. and Pan, J. (2002). A comparison of epistemological beliefs and learn­ing from science text between American and Chinese high school students. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychol-ogy of beliefs about knowledge and knowing (365–86). Mahwah, NJ: Law­rence Erlbaum Associates.

Root­Bernstein, R. (1984). On defining a scientific theory: Creationism considered. In A. Montagu (Ed.), Science and creationism (69–94). New York: Oxford University Press.

Sandoval, W. A. (2005). Understanding students’ practical epistemologies and their influence on learning through inquiry. Science Education, 89, 634–56.

Schommer­Aikins, M. (1993). Epistemological development and academic performance among secondary students. Journal of Educational Psychol-ogy, 85, 406–11.

(2002). An evolving theoretical framework for an epistemological belief sys­tem. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (103–8). Mahwah, NJ: Lawrence Erlbaum Associates.

(2004). Explaining the epistemological belief system: Introducing the embedded systemic model and coordinated research approach. Educa-tional Psychologist, 39, 19–29.

Schommer­Aikins, M., Duell, O. K., and Hutter, R. (2005). Epistemological beliefs, mathematical problem­solving beliefs, and academic performance of middle school students. Elementary School Journal, 105(3), 289.

Seigel, H. (1988). Educating reason: Rationality, critical thinking, and education. London: Routledge and Kegan Paul.

Simmons, M. and Zeidker, D. L. (2003). Beliefs in the nature of science and responses to socioscientific issues. In D. L. Zeidler (Ed.), The role of moral reasoning on socioscientific issues and discourse in science education. The Neth­erland: Kuwer Academic Press.

Songer, N. B. (1989). Promoting integration of instructed and natural world knowl-edge in thermodynamics. Unpublished doctoral dissertation, University of California, Berkeley, CA.

Toulmin, S. (1958). The uses of argument. New York: Cambridge University Press.

Tsai, C.­C. (1998). An analysis of scientific epistemological beliefs and learning orientations of Taiwanese eighth graders. Science Education, 82, 473–89.

(2000). Relationships between student scientific epistemological beliefs and perceptions of constructivist learning environments. Educational Research, 42, 193–205.

(2004). Beyond cognitive and metacognitive tools: The use of the Internet as an “epistemological” tool for instruction. British Journal of Educational Technology, 35, 525–36.

Voss, J. F., Perkins, D. N., and Segal, J. W. (Eds.) (1991). Informal reasoning and education. Hillsdale, NJ: Lawrence Erlbaum Associates.

Fang­Ying Yang and Chin­Chung Tsai 157

Vygotsky, L. S. (1978). Mind in society. Cambridge: Harvard University Press.Wellman, H. M., Cross, D., and Watson, J. (2001). Meta­analysis of theory­

of­mind development: The truth about false belief. Child Development, 72, 655–84.

Willis, S. and Schaie, K. W. (1993). Everyday cognition: Taxonomic and meth­odological considerations. In J. M Puckett and H. W. Reese (Eds.), Mech-anisms of everyday cognition (33–53). Hillsdale, NJ: Lawrence Erlbaum Associates.

Yang, F. Y. (2004). Exploring high school students’ use of theory and evidence in an everyday context: The role of scientific thinking in environmental science decision­making. International Journal of Science Education, 26, 1345–64.

(2005). Student views concerning evidence and the expert in reasoning a socio­scientific issue and personal epistemology. Educational Studies, 31, 65–84.

Yang, F. Y. and Anderson, O. R. (2003). Senior high school students’ prefer­ence and reasoning modes about nuclear energy use. International Journal of Science Education, 25, 221–44.

Yang, F. Y., Tsai, C.­C., and Lee, J.­S. (2005). The mode of informal reasoning and the personal epistemology of the 6th graders. Paper presented in the 11th Biennial Conference European Association for Research on Learning and Instruction (EARLI), Nicosia, Cyprus, August.

Yang, F. Y. and Tsai, C.­C. (in press). Reasoning on science­related uncer­tain issues and epistemological perspectives among children. Instructional Science.

Zimmerman, C. (2000). The development of scientific reasoning skills. Devel-opmental Review, 20, 99–149.

A PPEN DI X ASTOR IES A N D OPEN­EN DED SURV EY QU EST IONS USED TO ASSESS SCIEN T IF IC R EASON I NG A N D PER SONA L EPIST EMOLOGY I N ST U DY 2

(a) Basic story of the target issue:

Due to a shortage of water supply, the government planned to drill wells in an area full of underground water. However, the residents rejected such a project because a previous ground subsidence incident had occurred during a major earthquake. The residents were afraid that the excessive use of underground water would induce further ground subsidence. Questions were then posted for students to jus­tify residents’ action and information released by the water company. The questions used and the respective coding targets are presented in Appendix D.

An epistemic framework for scientific reasoning158

(b) Open-ended questions:

(1) What do you think caused the previous ground subsidence in the town? Why?

(2) Do you think the residents’ resistance was reasonable? Why?(3) Do you believe the claim made by the water company that they had

done careful investigation before the decision of well­drilling was made? Why?

What could they do to make you believe?•(4) Are you sure about your answers? Why?

What information would you need to help make better judgments?•

A PPEN DI X BSTOR IES A N D OPEN­EN DED SURV EY QU EST IONS USED TO ASSESS SCIEN T IF IC R EASON I NG A N D PER SONA L EPIST EMOLOGY I N ST U DY 3

(a) Basic story of the target issue:

The issue in discussion was about a protest event launched by some construction companies against government policy banning house con­struction close to a river band in a city. The government banned the construction because it was suspected that overloaded construction had caused floods in the rainy seasons. However, construction companies claimed that they would not accept the ban unless some “ scientific or quantitative data” were presented to prove the correlation between flood disasters and house constructions. Questions regarding the forms of sci­entific evidence and the role that expert opinions played were presented to probe personal views toward the nature of knowing in science.

(b) The open-ended questions:

(1) What do you think caused the previous ground subsidence in the town? Why?

(2) Do you think the residents’ resistance was reasonable? Why?(3) Do you believe the claim made by the water company that they had

done careful investigation before the decision of well­drilling was made? Why?

What could they do to make you believe?•(4) Are you sure about your answers? Why?

What information would you need to help make better judgments?•

Fang­Ying Yang and Chin­Chung Tsai 159

A PPEN DI X CSTOR IES A N D OPEN­EN DED SURV EY QU EST IONS USED TO ASSESS SCIEN T IF IC R EASON I NG A N D PER SONA L EPIST EMOLOGY I N ST U DY 4

(a) Basic story of the earthquake issue:

The earthquake prediction is a pure­science and domain­specific issue which involved two different expert groups arguing whether the earth­quake is predictable. One group of experts, by observing the abnormal behavior of some certain fish species, believed that they have found a way to predict earthquakes. On the other hand, by referring to the fail­ure of an astronomic experiment which was recognized as a scientific way to predict earthquakes, the other group of experts insisted that the mechanism of earthquake is too complicated to be understood. There­fore, prediction is not possible. The land subsidence issue as previously described is multidimensional by nature in that it concerned not only the scientific problems regarding the geological investigations but also social problems about the loss of human lives and properties.

(b) The interview questions:

Phase I: Prior to the reading of any news reports

(1) Do you think earthquakes can be predicted?(2) Do you think that scientists or experts have the same opinions when

considering the prediction issue? Why?(3) When doing scientific research, do you think experts would reach

an agreement eventually? Why?

Phase II: Proceed to the reading of Issue I: Earthquake prediction

(4) What is the difference between the two experiments?(5) Do you now think that earthquakes can be predicted? Why?(6) Which news do you believe more? Why?(7) What could the experiment you do not favor do to make you

believe? Please explain it.

Phase III: Proceed to the reading of Issue II: Land subsidence

(8) What was the cause of land subsidence reported in the news?(9) Do you think the protest by residents is reasonable? Why?

An epistemic framework for scientific reasoning160

Coding target

Examples of open­ended questions

Examples of responses

Analysis

Assessing beliefs about the nature of knowledge (certainty of knowledge)

(Earthquake issue) (2) Do you think that scientists or experts have the same opinions regarding the issue of earthquake predictions? Why?

(30124) No, everyone has his/her own thought. (31525) No, people think differently.

30124 – Multiplist 31525 – Absolutist

(3) Do you think that experts would reach an agreement eventually when doing any scientific studies? Why?

(30124) No, everyone has his/her own thought. (31525) Yes, because they are observing the same thing.

Assessing prior belief/theory

(5) Do you think that earthquakes are predictable?

(30124) No. (31525) No.

(10) The water company claimed the safety of the well­drilling. Do you believe their claim? Why?

(11) Are you sure about your ideas with respect to above questions? Why?

(12) What do you think the water company should do to make the resi­dents believe their claim? Please explain it.

Note. Coding standards – Students’ responses regarding the certainty of knowledge (such as questions 2 and 3) were cross­compared to Kuhn’s and Perry’s concepts about the nature of knowledge, while responses that involved expert and evidence (such as questions 5, 6, 9, and 10) were cross­checked with the concepts about justification described in the reflective judgment model. In addition, student’s reflections on their own thoughts (question 11) were examined based on Kuhn’s ideas about the necessity for critical thinking.

A PPEN DI X DEX A M PL ES OF CODI NG A NA LYSIS FOR ST U DEN T R ESPONSES I N ST U DY 5

Fang­Ying Yang and Chin­Chung Tsai 161

Appendix D (cont.)

Coding target

Examples of open­ended questions

Examples of responses

Analysis

Assessing beliefs about the process of knowing and judgmental criteria

(After reading a news report about earthquake predictions) (6) Do you think now that earthquakes can be predicted? Why?

(30124) Yes, fish have the ability to make predictions. (31525) No, many scientific methods failed to predict earthquakes.

Views toward expert opinions: 30124 – (Change of personal theory) Expert opinions justify personal theory. 30525 – (No change of personal theory) Expert opinions defend/confirm personal theory. Judgmental criteria:30124 – The validity of study.31525 – The consistency to personal belief/theory.

(7) Which news do you believe more? Why?

(30124) The astronomy study does not look like a real study. (30525) I think that earthquakes are unpredictable … the astronomy study confirmed my thought.

Assessing coordination of theory and evidence

(8) What could the expert group which you do not believe do to make you believe their claim? Please explain it.

(30124) The astronomy group needs to do the study more carefully. (31525) If the fish group does the study again and obtains the same result, I will believe their claim.

30124 – No coordination of theory and evidence. 31525 – Coordination of theory and evidence.

Assessing beliefs about the process of knowing and judgmental criteria

(Land-subsidence issue) (9) Do you think that the residents’ protest against the well­drilling project is reasonable? Why?

(30124) Yes, when an earthquake comes, the land will subside if there are many wells. (31525) Yes, the residents were protecting themselves.

Views toward expert claims: 30124 – Expert claims are assertions that could be wrong. 30525 – A wrong assertion. Judgmental criteria: 30124 – Past experiences. 31525 – Personal safety, company benefits (social aspect of science).

(10) The water company claimed the safety of the well­drilling project. Do you believe their claim? Why?

(30124) No, the land had subsided before, so the water company’s claim cannot be trusted.

An epistemic framework for scientific reasoning162

Coding target

Examples of open­ended questions

Examples of responses

Analysis

(31525) No, the water company considered only their own benefits.

Assessing reflective reasoning

(11) Are you sure about your ideas with respect to above questions? Why?

(30124) Yes, because I’ve read the reports carefully. (31525) Yes, the water company just wanted to realize their plan.

30124 – Reflective reasoning to assure one’s own thought. 31525 – Reflective reasoning to assure one’s own thought.

Assessing coordination of theory and evidence

(12) What do you think the water company should do to make the residents believe their claim? Please explain it.

(30124) The water company should dig the well elsewhere. (31525) Stop the project.

30124 – No coordination of theory and evidence. 31525 – No coordination of theory and evidence.

Appendix D (cont.)

163

6 Who knows what and who can we believe? Epistemological beliefs are beliefs about knowledge (mostly) to be attained from others

Rainer Bromme, Dorothe Kienhues, and Torsten PorschUniversity of Muenster

Introduction

In this chapter we will discuss some implications of the distributed nature of knowledge for the study and for the teaching of epistemo­logical beliefs. We will elaborate on the following argument: due to the division of labor in modern societies, knowledge is distributed and used unevenly. Most knowledge claims are based on specialized knowledge provided by specialized experts, and the knowledge is organized into disciplines, reflecting such specialization. In the following chapter, the uneven distribution and use of knowledge is called “division of cognitive labor.” Because much of our knowledge is acquired from others, most everyday epistemological issues regard the assessment of knowledge claims made by others who are experts for issues we are not experts for. Improving epistemological judgments, therefore, requires the improve­ment of a person’s capacity to understand how specialized knowledge is distributed (who knows what) and to evaluate expert sources (whom to believe). In the first part of this chapter the notion of the “division of cognitive labor” is explained. Secondly, how the uneven distribution of cognitive labor and the dependency from experts has been conceived in research on epistemological beliefs will be discussed. The well­known dimensions “source” and “justification of knowledge” will take the sec­tion’s center stage. The dimension “source” describes knowledge which is not constructed by the knower and implies a rather negative view on experts’ knowledge. The dimension “justification of knowledge” refers to beliefs necessary for the assessment of experts’ knowledge claims, but the “sophisticated” subject is mostly conceived of as overcoming the division of cognitive labor.

Who knows what and who can we believe? 164

Most problems laypersons are confronted with in their everyday lives require knowledge and expertise which goes far beyond laypersons’ understanding. Therefore, we will question the normative assumption that one’s own knowledge is better than knowledge attained from others which underlies many research approaches on epistemological beliefs. We will then provide some empirical evidence from developmental psy­chology about children’s intuitive understanding of the division of cog­nitive labor and we will ask how schooling fosters and impedes such understanding. We also report on a study about primary school stu­dents’ assessment of knowledge sources. It will suggest that references for sources depend on subject matter contexts as well as on school tasks themselves. Furthermore, the study will offer a methodical approach to the measurement of competencies in the recognition of, and dealings with, divided knowledge. We exemplify what kinds of capacities should be fostered when it comes to the improvement of epistemological beliefs in schools. Finally, we will discuss implications for teaching and for research.

The division of cognitive labor: the uneven distribution of knowledge within a society requires a continuous assessment of knowledge claims

Predominantly, our knowledge acquired during life does not derive from personal, direct experience, but from other people’s experiences, con­siderations, or analyses. At first it comes from parents and peers, later from teachers, and it is mostly transferred through cultural artifacts like books, TV, or the Internet (Bergstrom, Moehlmann, and Boyer, 2006). In acquiring such knowledge throughout informal as well as for­mal education, we acquire a share of the knowledge distributed within a society. In the following section, we will refer to this knowledge as distributed knowledge. Acquisition of knowledge from others starts very early. Developmental psychology has shown that even preschoolers have impressively clear concepts about several topics which are beyond their personal experiences. For example, they have realistic ideas about how babies grow, how mind and behavior are related, and what makes up illnesses (Goswami, 2008). “It is unlikely that children begin to under­stand any of these domains – which involve causal processes hidden from their view – unless they are given relevant information by adults” (Harris, 2001, p. 498).

In the Piagetian tradition children’s cognitive development has been conceived of as being mostly driven by a child’s personal, first­hand experience with her or his environment. Paul Harris (2001)

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 165

has challenged the widespread view of a young child as a “stubborn autodidact”: “Piaget ascribed virtually no educative role to the testi­mony that adults might offer, to the extent that children might incor­porate such teaching into their cognitive repertoire, Piaget dismissed it as mere ‘verbal’ knowledge, rather than genuine understanding” ( Harris, 2001, p. 495). In contrast, Harris concluded from his and his colleagues’ experiments that even preschoolers are willing and able to reason from premises (all cats bark) which they learned from others (cats from another planets do bark), even when such premises contra­dict their own first­hand experiences.

The individual’s capacity to reflect critically about knowledge claims is crucial because the reliance on the knowledge of others does not end with childhood. Due to the division of labor within modern societies such dependency on others remains throughout an adult’s whole life. Advances in science and technology have led – at least in the industrial­ized nations – to an enormous growth of scientific knowledge simulta­neously accompanied by specialization and differentiation. One of the key features of modern societies is the uneven distribution of knowl­edge across its members (Stehr, 1994). The notion that to be a true polymath seems impossible in our times (Keil et al., 2008) indicates the unstoppable growth of content and complexity in almost all fields of knowledge.

Knowledge does not only develop rapidly; it is also heavily specialized which will be conceived of in the following as uneven division of cognitive labor. Not only children but also adults remain laypersons throughout their whole lifetime with regard to most topics and domains of knowl­edge available in a society. We are laypersons with regard to knowledge that is relevant for our own life as a citizen: “For example, the right to vote for a representative in the government requires the ability to evalu­ate the views and beliefs of candidates and/or parties before making an educated choice” (Haerle and Bendixen, 2008, p. 172). As a profes­sional, the division of cognitive labor arises from the increasing division of labor in the working world and in our private lives. That implies also a continuous dependency on experts of all disciplines. It is not possible to study medicine in the case of an illness or to become a lawyer when a legal problem occurs in one’s life course. The uneven distribution of knowledge within a society requires the continuous assessment of knowledge claims provided by all kinds of experts.

The need for a critical assessment of knowledge claims is reinforced even further through modern information technologies. For example, lay­persons (non­experts) can now search for information on the Internet at any time and do not only come across scientific evidence that has already

Who knows what and who can we believe? 166

been prepared for general consumption, but also access information actually intended only for the discourse within science itself.

Proceeding from a specific problem (prototypical examples are web searches on health­related lifestyle information or the search for a “sec­ond” opinion on the diagnosis and treatment of a specific disease), it is very easy to obtain a host of scientific or science­based information (Bromme et al., 2005). This increases also the availability of fragile and contradictory evidence along with the need to interpret this evi­dence and evaluate how useful it may be for solving the problem at hand (Irwin and Wayne, 1996).

The strong public interest in science programs on television, in the science sections of newspapers, as well as in specialized popular sci­ence journals, indicates how much the general public perceives this need. The various providers in this market also contribute to ensur­ing that the availability of science­based knowledge is unproblematic. However, it is the interpretation of this evidence that becomes a major challenge. Empirical research has shown that when laypersons stumble across conflicting evidence on the Internet, they have some difficulty in performing such evaluation processes (Bråten et al., 2005; Eysenbach and Köhler, 2002; Mason and Boldrin, 2008; Tsai and Chuang, 2005; Whitmire, 2004; Wu and Tsai, 2007), as long as they are not supported instructionally (Stadtler and Bromme, 2007).

Several studies have shown the impact of epistemological beliefs on such evaluation processes. In a recent study, Strømsø et al. (2008) describe negative correlations between beliefs about the source of knowl­edge and the assessment of the trustworthiness of texts, indicating that students trust more in sources if they rely more on external authority. Furthermore, justification for knowing beliefs predicted the ability to differentiate between the trustworthiness of different genres (science texts versus newspapers): students noticing knowledge as constructed trusted more in science texts. Simplicity of knowledge beliefs correlated negatively with trustworthiness ratings. Students with a more theoreti­cal and complex view of knowledge trusted a newspaper article less.

Whitmire (2004) found that participants’ level of epistemological beliefs (referring to Baxter Magolda’s (1992) epistemological reflection model) affects information­seeking behavior. For example, participants holding a less advanced, absolutistic view predominately selected infor­mation sources consistent with their views, avoiding conflicting infor­mation. Furthermore, they were not able to determine the authority or usefulness of conflicting sources. In contrast, participants with more advanced beliefs (“transitional believers”) knew several ways to evaluate

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 167

web sources. For example, they examined the source of a website or the institutional affiliation of a website author. They additionally were interested in conflicting information.

How has the uneven distribution of cognitive labor and the dependency on experts been conceived of in research on epistemological beliefs?

Since the beginning of research on epistemological beliefs, asking sub­jects what they think about knowledge claims provided by authorities or experts has been a prominent testing probe for scrutinizing such beliefs. For example, Kuhn (1991) asked her subjects “Do experts know for sure what causes___?” and “Would it be possible for experts to find out for sure if they studied this problem long and carefully enough?” (pp. 172–3). She presumes that “The questions about expertise in fact prove the most revealing of all the epistemological questions and form the basis for the category system represented in this chapter” (Kuhn, 1991, p. 172). Her category system describes epistemological theories as sets of propositions about personal certainty (among other issues), which include several assumptions about experts’ knowledge. For example, subjects with absolutist epistemological theories claimed that experts know for sure because of their experience and because “They have a bunch of facts” (p. 174). Multiplists showed a very skeptical view of expertise, up to a denial of the existence of experts: “Nobody is really an expert. They always have room for improvement and learning. I don’t think there are anything like experts” (p. 178). Subjects with eval­uative theories stated, for example, “I think I’d be sure if I did a lot of research like the experts do. Right now, they’re probably more correct than I” (p. 189).

The widely used definition of personal epistemology as beliefs with four identifiable and more­or­less interrelated dimensions (Hofer and Pintrich, 1997) refers to expert knowledge in two of the four dimensions. The first two dimensions represent the nature of knowl­edge: (1) the certainty of knowledge is focused on the perceived sta­bility and the strength of supporting evidence, and (2) the structure of knowledge describes the relative connectedness of knowledge. The remaining two dimensions describe the nature of knowing: (3) the justification of knowledge explains how individuals proceed to evalu­ate and warrant knowledge claims, and (4) the source of knowledge describes where knowledge resides, internally and/or externally. The latter two dimensions refer to expert knowledge and they do so in a specific and, as we will explain further, problematic way.

Who knows what and who can we believe? 168

The dimension ‘‘source” implies a rather negative view of experts’ knowledge because it is knowledge from the “outside,” not constructed by the knower.

The underlying normative assumption of the previously mentioned dimension “source” states that knowledge that is developed by the knower herself/himself is better than knowledge that comes from the outside.

At lower levels of most of the models, knowledge originates outside the self and resides in external authority, from whom it may be transmitted. The evolving conception of self as knower, with the ability to construct knowledge in interaction with others, is a developmental turning point of most models reviewed. Perry (1970) described this awareness as one of the shifts in his model, when ‘‘the person, previously a holder of mean­ing, becomes a maker of meaning” (p. 87) (Hofer, 2000, p. 381).

The dimension source refers to “authorities” as the outside sources of knowledge. This is not confined to authority which is based on the uneven distribution of knowledge; it could also be an authority of politi­cal, economic, or religious power. Nevertheless, this dimension conveys the sometimes implicit appreciation of knowledge generated by oneself as being “better” than knowledge attained from the outside (but see our later description of Kuhn’s (1991) evaluativist, who has a more bal­anced view). This normative assumption might be due to the Piagetian legacy within research programs on epistemological beliefs and it is also due to the prevailing normative ideal of a student as an autonomous learner who makes up her/his mind independently from authorities. This ideal goes back to Perry’s (1970) seminal work from which much of the recent research on personal epistemological beliefs emerged.

The dimension “justification of knowledge” refers to beliefs necessary for the assessment of experts’ knowledge claims, but the “sophisticated” subject is conceived of as overcoming the division of cognitive labor.

This dimension includes how individuals evaluate knowledge claims, including the use of evidence; the use they make of authority and exper­tise; and their evaluation of experts. In the reflective judgment model (King and Kitchener, 1994), individuals at lower levels justify beliefs through observation or authority, or on the basis of what feels right, when knowledge is uncertain. Only at higher stages do individuals use rules of inquiry and begin to personally evaluate and integrate the views of experts (Hofer, 2000, p. 381).

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 169

This dimension clearly refers to the assessment of knowledge claims and – as was explained previously – experts’ knowledge claims are a prominent example when it comes to data collection about beliefs of this dimension. So far, this dimension allows for the uneven division of cognitive labor. Nevertheless, in line with the implicit assumption underlying the source dimension, the sophisticated subject is conceived of as doing such assessment of knowledge claims in her/his own mind, in a sense, like the expert. The implicit underlying normative assump­tion strives for the subjects’ capability to overcome the division of cog­nitive labor (but see below for a discussion of “evaluativist” beliefs). It is the implicit ideal of the subject as a problem­solver who asks sev­eral experts, understands their claims, thinks about them, and finally comes up with her/his opinion about the issue at stake, so to say, on par with the expert.

For example, in a study on primary school teachers’ beliefs about knowing, Brownlee (2003) divided the comments made by the par­ticipants regarding their epistemological beliefs into four main cat­egories. Participants assigned to the lowest category, called received beliefs, assume experts to facilitate the reception of absolute truth, while participants belonging to the highest category, called construc­tivist beliefs, view experts as facilitating the construction of reasoned truth. Especially among researchers on epistemological beliefs of school students it is common to emphasize the function of experts’ opinion as “footholds” (Hammer and Elby, 2003) for students’ construction of their own understanding. In doing so, students become well­equipped to deal with school knowledge, but how are they prepared when con­fronted (for example, on the Internet) with knowledge claims which are beyond the scope of the curriculum taught in schools?

What is true? Who to believe? Two questions that should be kept apart

Summarizing the underlying assumptions of the prevailing four­ dimension model we suggest a conceptual distinction between first-hand evaluation and second-hand evaluation of knowledge claims, borrowing the notion of first-hand experience from the science educa­tion literature.

First-hand evaluation: what is true? The veracity of a knowledge claim could be assessed directly, for example by thinking about the consist­ency of the claims made, by comparing it with other pieces of knowledge, or by thinking critically about its logical coherence and cohesiveness. Such first-hand evaluation processes are the focus of most attempts to

Who knows what and who can we believe? 170

improve epistemological beliefs and they are also at the heart of what is called and taught as “critical thinking.”

Second-hand evaluation: who to believe? Unfortunately, first­hand evaluations require a lot of domain­specific expertise and knowledge. Many problems laypersons are confronted with in their everyday private lives or as citizens participating in political decisions require knowledge and expertise that goes far beyond their understanding. Examples are medical, legal, and economic decisions where laypersons not only have to rely on experts’ explanations and advice, but where they also have to cope with the fact that they are not able to acquire the knowledge and skills which would be necessary to assess the experts’ knowledge claims. This problem has become evident, for example, in the context of shared decision­making approaches in medicine, where patients are asked to make an informed decision on their own treatment (Runde et al., 2007; Tuckett et al., 1985). Nowadays it is no problem for patients to get access to detailed medical knowledge via the Internet. But the understanding of many knowledge claims still requires conceptual knowledge about medical, biological, and/or chemical structures and processes which goes far beyond a non­expert’s understanding of medi­cine, chemistry, and biology.

First­hand judgments about the relevance and veracity of knowledge claims require domain­specific knowledge and skills. They must be based on topic­related and on ontological knowledge about the realm of reality the knowledge claims refer to (Bromme et al., 2008). Therefore, in many cases a second-hand approach is necessary. The question “Which knowledge claim is true and relevant?” often must be transformed into the question “Which source of knowledge is credible and relevant?” This question is logically different from the “What is true?” question that asks for the person’s own, so to say, direct judgment of truth and which is the focus of most research on epistemological beliefs. This question is psy-chologically different in terms of the reasoning processes and knowledge structures required to answer it. Sophisticated epistemological judg­ments within second-hand evaluation require knowledge about the divi­sion of cognitive labor: who is “responsible” for which topic and how do we decide about the credibility and relevance of different and sometimes competing sources of knowledge claims?

It must be emphasized that the conceptual distinction between first­ and second­hand evaluations of knowledge claims is made here for the sake of analytical clarity. Practically, both questions are often mixed and many of the decisions described above, where laypersons are dependent on experts’ knowledge, do require a balance between the first­hand and second­hand evaluations. In other words, they require a

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 171

direct, personal assessment of the veracity of a knowledge claim (could that claim be true, does it fit with my experience?) as well as a second­hand assessment (what are the reasons to believe that expert source?). In the same vein, several authors have conceived of such balanced judg­ments as the most developed (sophisticated) judgments, based on eval­uativistic epistemological beliefs. For example, Kuhn (1991) describes the highest mastery level of argumentation skills as based on evaluative theories which enclose such a balanced view of experts:

Evaluative epistemologists thus share the view that the greater observation, examination and analysis of an issue that an expert undertakes mean that greater certainty is attached to the experts’ view than to the average person’s. Though absolute certainty is impossible, experts may come closer to achieving such certainty than nonexperts. (p. 189)

A balanced relation between knowledge generated by the knower and knowledge attained from others has been described as the most advanced level of epistemological thinking.

Evaluativistic thinking describes a way of knowing that focuses on evaluation and decision­making among differing viewpoints. In the developmental frame­works in the field of personal epistemology this is often defined as evaluativ­ism – the most advanced level of epistemic development (Hofer and Pintrich, 1997; King and Kitchener, 1994; Kuhn et al., 2000). Because evaluativistic thinking integrates subjective and objective views of knowledge and considers its complexity and uncertainty in relation to its context (Kuhn and Weinstock, 2002) we argue that this form of thinking is most evocative for, and expression of, the idea of democracy and what makes a “good” citizen in Western cultures. (Haerle and Bendixen, 2008, p. 167)

Note that this is a normative description of an epistemological stance which, as Kuhn (1991) reports, has been adopted only by a few of her research subjects. The majority (adopting absolutist and multiplist views) have other views about the credibility and role of experts (as described previously). Recent research on epistemological beliefs has described those views always with a focus on the tension between one’s own knowledge and knowledge attained from others.

It is psychologically implausible to assume that only the minority of people who have adopted an evaluativistic view have the capacity to deal with the division of cognitive labor in a way that allows for appro­priate assessments of knowledge claims provided by experts. Due to the division of labor, everyone is a layperson with regard to most knowledge domains. Most subjects will not be able to fully understand the knowl­edge they attain from experts, but nevertheless they have to deal with it throughout their whole adult life. Therefore, they must have some

Who knows what and who can we believe? 172

capacities (beliefs, knowledge, skills) for “second­hand evaluation” of knowledge claims. In the following section, we will discuss the compo­nents of such capacities and we will mention some research in order to exemplify how the notion of epistemological beliefs and epistemologi­cal thinking could be extended in order to describe such capacities. In order to rely on experts’ knowledge, one needs an understanding of the distribution of knowledge structures in society as well as different sub­stantial skills to access and evaluate adequate information.

Who knows what? Understanding the division of cognitive labor

What do students and adults know and believe about real experts? How do they conceive of the social distribution of expertise, its organization in professions, academic domains, and social institutions? Due to the prevailing research perspective on beliefs about the balance between subjective and objective forms of knowledge, there are large blind spots on the research maps which have been drawn about the landscapes of epistemological beliefs so far.1

Understanding distributed knowledge structures. First, one must know who knows what (i.e., how knowledge structures are organized within a society). At least for adults and for most topics, the answer to the question “Who knows what?” seems to be quite obvious at first glance. Knowledge about broken cars can be expected from mechanics and knowledge about health problems comes from medical doctors. But there are many topics where the picture is less clear: who is the best to ask about the relationship between the consumption of butter and heart disease? It could be medical doctors, nutrition specialists, or biologists. Who knows how to fix a breakdown of the Internet access of a personal computer: A technician of the manufacturer, a student of informatics, or a neighbor who is a computer nerd?

Understanding the distribution of knowledge structures includes an understanding of conceptual structures as well as of the distribu­tion of expertise within a society. The prevailing pattern of knowledge

1 We do not contend that the social origins of epistemological beliefs in general have been neglected within recent research on epistemological beliefs. Several authors have emphasized the cultural (see, for example, Hofer, 2008), contextual (Limon, 2006), and socio­cultural embeddedness (Buehl and Alexander, 2006; Hammer and Elby, 2003; Rozendaal et al., 2001) of personal epistemological beliefs. Starting from the conceptual distinction of personal beliefs about knowledge and the knowledge such beliefs refer to, the theoretical notion of “division of cognitive labor” refers to the social distribution and to the function of knowledge within a society (for example, the distinction between expert knowledge and everyday knowledge).

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 173

distribution follows the academic disciplines (beginning with the fun­damental distinction between humanities and sciences, which are com­binations of disciplines like history, philosophy, etc., and, respectively, biology, physics, medicine, etc.), but this pattern is not unequivocal with regard to the distribution of expertise. If, for example, a layperson seek­ing nutrition information on the Internet comes across the claim Eating butter leads to high cholesterol, which might lead to cardiac infarction, the question arises as to who has the expertise to corroborate such a knowl­edge claim? It is a medical topic, but it is also a biological and a chemical topic and, furthermore, a topic for nutrition specialists. It is necessary to have some amount of knowledge about a certain issue in order to figure out who might know what about it (Bromme et al., 2008).

Developmental psychology evidence about children’s basic understanding of the distribution of cognitive labor

In order to put a specific issue into the whole personal picture of une­venly distributed knowledge, it is necessary to have a general idea about how knowledge is clustered within a society. Young children already should show a (basic) understanding of knowledge clusters. There is some evidence (recently published by Frank Keil and his colleagues; see Keil et al. (2008) for an overview) that even young children are aware that people have different areas of expertise, and that they know how knowledge is clustered in the minds of others.

For instance, in a study by Lutz and Keil (2002), fifty­six children of different ages (three­, four­, and five­year­olds) had to judge which of two experts (a car mechanic or a doctor) would know more about a specific topic. The topics asked for three different types of expert knowledge (observable knowledge, knowledge about functioning, and knowledge of underlying scientific principles). For each of the pro­fessions, four questions focused on the observable knowledge a car mechanic or a doctor uses (“stereotypical role”), four questions focused on knowledge about the functioning of machines or the functioning of people (“normal functioning”), and four questions focused on knowl­edge of the scientific principles underlying the different domains of expertise (“underlying principles”). In sum, twenty­four questions in the form of “Who would know more about ____ (a specific topic)?” were asked. Results showed that the four­ and five­year­olds were able to attribute knowledge correctly for all three item types, whereas the three­year­olds performed better than chance merely on the stereotypi­cal role items. Children have an early notion of the division of cognitive labor, as they assume that different experts know different things. This

Who knows what and who can we believe? 174

understanding seems to be limited to observable knowledge at first, but then children begin to understand that expertise comprises more than only the observable knowledge and start to cluster information into distinct domains (at least as long as these are familiar domains of expertise, Lutz and Keil, 2002).

Another one of their studies focused on children’s clustering of knowl­edge and their ability to infer another person’s knowledge ( Danovitch and Keil, 2004). In the first experiment, children of different ages (six­, eight­, ten­, eleven­, and twelve­year­olds) had to decide about which of two specific topics an expert (introduced in an initial statement) would know more. An exemplary initial statement on the topic someone knows a lot about is “This person knows all about why keys don’t work as well if they are old and worn out.” The topics to choose between related to the topic provided in this initial statement in different ways. The first option to choose included the same discipline (either physics or social psychology) but a different topic and a different goal (“Do they know more about why it’s hard to turn the wheels on a bicycle if they are rusty?”). The other option involved either a different discipline but the same topic and the same goal (“Or, do they know more about why people sometimes forget which key opens the car door or the trunk?”), or it involved the same topic but a different goal and discipline (“Or, do they know more about why keys were first used in ancient Rome?”). All items provided were built in the same format and the presentation of the two options was counterbalanced. Results reveal that clustering knowledge according to disciplines (rather than according to topics) increases with age and develops gradually. However, some of the oldest children in the sample maintained clustering knowledge according to topics, so this development is not considered as finished by the age of twelve. A subsequent experiment (Danovitch and Keil, 2004) strength­ens the findings of the first experiment.

In sum, the study of Danovitch and Keil (2004) shows that children’s idea of how knowledge is clustered in the minds of others changes dur­ing elementary school. In contrast to younger children who predomi­nately generalize another person’s knowledge due to topics and goals, older children (around ten years of age) consider more principles related to disciplines. Nevertheless, it is worth emphasizing that the youngest children in the study already show an understanding of knowledge clus­ters. Danovitch and Keil (2004) conclude that “a very early emerging and compelling sense of the importance of knowing who knows what may drive the need to understand further the division of cognitive labor and thereby cause the pattern of development found” (p. 928).

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 175

Through subsequent studies, Keil and his colleagues (Keil et al., 2008) substantiate that five­year­olds can already fragilely cluster knowledge in a discipline­based way, although they are not assumed to have an explicit awareness of these different disciplines, and this way of clustering still competes with other ways.

In our view, the developmental psychology work cited previously is seminal for future research on epistemological beliefs.2 For example, it exemplifies how the structure of knowledge dimension (traditionally understood only as describing the relative connectedness of knowledge) could be conceived of and researched in a new and fruitful way. Such further research is needed because we do not know how students in higher grades and how adults conceive of the distribution of knowledge and how they conceive of the division of cognitive labor.

Is an understanding of the division of cognitive labor supported in schools?

In the following section we will discuss how schooling might foster or impede understanding the division of cognitive labor in a society. While the previously mentioned studies give some evidence that a basic under­standing of discipline structures increases with the age of children, one cannot conclude that schooling is directly helpful for developing an understanding of how knowledge is structured in the world. Most of the school curriculum is organized according disciplines, but that does not mean that schooling automatically supports children’s understanding of knowledge structures.

Traditional school subjects (mathematics, science, history, sport, lit­eracy) structure the whole of school life. They do not only define the content of the curriculum, but also most of its organizational structures. According to different school subjects, students move from room to room, they interact with different teachers specialized according to such disciplines, and the course of time is structured in units which are more­or­less strictly defined with regard to these structures. Stodolsky (1988) has shown in a series of studies that even the ways of how interaction takes place in the classroom is impacted by the respective school subject.

2 Keil’s findings are in line with the evidence about children’s “theory of mind” (TOM, e.g., Liu et al., 2008; Wellman et al., 2001; Wimmer and Perner, 1983), but his research has a different focus. While TOM research scrutinizes the children’s understanding of the difference between their own minds and the minds of others, Keil’s approach focuses on children’s theories about the knowledge distribution among others (domain experts). Within this chapter we argue for an analogue extension of the focus of research on epistemological beliefs.

Who knows what and who can we believe? 176

Mathematics is taught in a different way than language studies. The ways of organizing as well as of evaluating students’ work differ between different school subjects, and students as well as teachers are well aware of such differences (Olafson and Schraw, 2006; Muis, 2004). Never­theless, that does not mean that students do understand the conceptual differences between school subjects, nor that they understand the con­ceptual differences between the corresponding academic disciplines.3

Stevens et al. (2005) have pointed to the lack of support for the devel­opment of a comparative understanding of school subjects. Students experience the school day as a sequence of time units, each assigned to a different conceptual understanding of the world, but the relationship between these understandings is not discussed. They realize that the same notions have a different meaning in different lessons (for exam­ple, explanation, discussion, and proof have different meanings in math­ematics, social science studies, and language studies) (Donald, 1995), but such relationships and differences are rarely discussed in school. “In other words, those least equipped to bring conceptual order to the school day – students themselves – end up shouldering the burden of having to do so” (Stevens et al., 2005, p. 127).

Even so­called “interdisciplinary” or “integrative” curricular acco­unts sometimes are not helpful as they focus on topics, not on disci­plines, neglecting the disciplinary differences instead of discussing them as competing knowledge assertions. Especially in primary schools, it is common to organize the curriculum in terms of topics instead of disciplines, in order to support bridging the gulf between school and children’s experiences out of school. Such educational accounts often are motivated by constructivistic or “problem­oriented” approaches, emphasizing the importance of the student’s personal, first­hand expe­riences. Such emphasis on students’ activities and perspectives for learning is well­founded by research (Bransford et al., 2000), but it must not necessarily be detrimental for teaching an understanding of disciplinary knowledge structures.

A study on source judgements of elementary school students in the context of different school subjects

Given that subject matter structures are fundamental organizing principles and, simultaneously, given that a comparative perspective

3 Note that school subjects in teaching and the corresponding academic disciplines are only weakly related to each other (Bromme, 1994; Schwab, 1987). Just as the contents to be learned in German lessons are not simplified German studies, but represent a canon of knowledge of their own, the contents of learning mathematics are not just simplifications of mathematics as it is taught in universities. The school subjects have a “life of their own” with their own logic; that is, the meaning of the concepts taught cannot be explained simply from the logic of the respective scientific disciplines.

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 177

on disciplines taught in schools is missing as outlined previously, it is an interesting open empirical question how students’ understanding of the structure of disciplines develops in schools. In order to exem­plify how this research question could be tackled, we will report in the following section on a study (Porsch et al., 2010) on primary school students (fourth­graders). This study approaches the question of how ideas about disciplines are related to different sources. So far, Keil and colleagues (2008) (see previous description) have already shown that even younger students have a basic understanding that knowl­edge in the minds of others is clustered in a discipline­based way. Sourcing, that is, identifying and remembering a source and its criti­cal characteristics when evaluating knowledge claims, is an important component of the capacity for second­hand evaluation. Therefore, in our study we additionally took into account students’ assessments of sources (e.g., books, Internet, and one’s own hands­on activities) for the acquisition of new information which could be relevant for solving a typical school task.

In a nutshell, we explored the impact of subject matter context (math and science lessons) curricular topic (e.g., why some objects sink and others float in water), and context on students’ assessments of sources (e.g., books, Internet, and one’s own hands­on activities) for the acquisition of new information which could be relevant for solv­ing a typical school task. By asking the students which source they would use in order to solve a word problem, the procedure we used was similar to the research method used by Frank Keil and his colleagues (2008) and is also similar to a procedure developed by Kalish (2002). These developmental psychologists asked for choices between differ­ent sources of knowledge (for example, different experts) in order to establish data about their subjects’ understanding of knowledge struc­tures. In doing so, they confound subjects’ ideas about sources and knowledge structures. Such a confounding might be innocuous as long as young children are studied who do not differentiate between knowl­edge structures and the persons which might represent them within a society. But, when it comes to school children it is itself an interesting empirical question how ideas about sources (who knows what best?) are related to ideas about knowledge structures. This is exactly what we have studied.

At first, we designed word problems which addressed three different curricular topics taught in German elementary schools in the fourth grade: (1) floating and sinking of objects in water; (2) sorting magni­tudes; and (3) assigning values. The appearance of tasks was adapted to the particular subject matter discipline (e.g., tasks in the math lesson involved more numbers). This enabled us to assemble word problems from very typical lessons and to maintain comparability of the logical

Who knows what and who can we believe? 178

operations required in order to solve the tasks. For example, in the cas­tle problem it is asked for the “magnitude of the castle” in the context of a science lesson and for its “length and broadness” in a math lesson context.

These six word problems (e.g., “Here you see several cubes. They are all equal in size, but have different weights. Some of them float when they are set into water, others sink. Find out from what weight the cubes sink”) were presented to two groups of students (see Table 6.1). One group received the problems within a mathematics lesson context; the second group received the problems in a Sachkunde lesson context.4

The word problems were presented in a questionnaire, but they did not need to be solved. Instead, the students were asked in which ways they would acquire knowledge “in order to find the right answer for sure.” For every word problem eight possible sources of knowing had to be rated on a five­point Likert scale (1 = totally wrong to 5 = totally true). These sources consisted of five external sources (parents, books, the Internet, teachers, and friends), and three internal sources of infor­mation by one’s own hands­on experience with the task (planning how to test a task solution and actually try it out as well as doing some cal­culations by themselves).

In a preliminary experiment with twenty­four children (twelve girls and twelve boys), we had successfully proven equality between the tasks with regard to task difficulty, comprehensibility, and school rel­evance. Furthermore, eight elementary school teachers rated the task independently.

Participants were 225 children (129 boys and 96 girls), from eleven German elementary schools. The questionnaire was administered in whole school classes. All children attended the fourth grade. Mean age was M = 9.6 (SD = .57) with a range from eight to eleven years. Girls (M = 9.49, SD = .54) were significantly younger than boys (M = 9.69, SD = .57) with t(223) = 2.56, p = .01. The children were randomly distributed to two independent samples of participants. In the first sample, 115 children (65 boys and 50 girls) worked with tasks from science lessons, and in the second, 110 children (64 boys and 46 girls) worked with tasks from math lessons. To avoid systematic interferences the splitting was also done within classes. The two samples did not differ in regard to age, school performance (measured by grades), and interest in school subjects.

4 In German elementary schools (first­ to fourth­grade) science as well as social study topics are taught within one subject matter domain, called “Sachkunde.”

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 179

At first, the previously mentioned a priori grouping of sources (external sources, internal sources) was checked by an explorative factor analysis confirming the assumed distinction between external sources (as teachers, parents, Internet), and internal sources (the stu­dents themselves) as a source of new insights. Furthermore, “calcu­lating” was conceived of as a separate source of one’s own hands­on experience.

Source differences between tasks depending on the interaction of subject matter context (comparison between students who had thought about the word problems in a math lesson context, and students who considered the word problems in a Sachkunde lesson), and curricular topic (comparisons within subjects) have been multivariately analyzed. A multiple analysis of variance (MANOVA) showed significant main effects for subject matter context (F(3,221) = 24.41, p < .01, g2

part = 0.25) and curricular topic (F(6,218) = 18.88, p < .01, g2

part = 0.34). Furthermore, a significant interaction between subject matter context and curricular topic was found (F(6,218) = 5.87, p < .01, g2

part = 0.14). External sources (books, friends, Internet, parents) were preferred more in the Sachkunde lesson context (see Figure 6.1), whereas calculating was preferred more in the math lesson context. Interestingly, there was no difference with regard to sources of one’s own hands­on experience (plan how to test a task solution and actually try out) and/or viewing the teacher as a source of knowing with regard to the subject matter context. But overall, internal sources (acquisition of knowledge by one’s own activities) were preferred more than external sources of knowledge (see Figure 6.2). To sum up, presenting tasks in the context of math

Table 6.1. Study design

Tasks presented in math lesson (N=110)

Tasks presented in Sachkunde lesson (N=115)

Curricular topic

Floating and sinking

Sorting magnitudes

Assign values

Floating and sinking

Sorting magnitudes

Assign values

Word problem context

Cube weight

Hundred board

Castle Cube weight

Hundred board

Castle

Stone in water

Bar chart Suit of armor

Stone in water

Bar chart Suit of armor

Who knows what and who can we believe? 180

lessons or Sachkunde lessons affected source judgement. Furthermore, certain sources were preferred in regard to different curricular topics.

Our results show that elementary school children have clear ideas about where to get information. They do prefer to get the insights necessary for their task solution by themselves (testing ideas, calculating) compared to external sources, but this difference is less strong in the Sachkunde context in comparison to the context of mathematics teaching. Given the logical similarities between the tasks presented within the Sachkunde and the mathematics contexts, the impact of the subject matter context is especially remarkable. Our study has clearly shown that even for fourth­graders, sources of knowledge and disciplines are not the same (and therefore should not be confounded), but that disciplines make up the background for the evaluation of sources, even if the logic of the tasks to be solved is the same between different subject matter contexts.

It can be concluded that preferences for sources depend on subject matter contexts as well as on the task itself. Students realize the require­ments of a school task itself, but they are also aware of the lesson’s frame given by subject matter contexts. Thus, one curricular topic implies different source preferences in math lessons as compared to Sachkunde lessons. Furthermore, this study offers a methodical approach to the measurement of competencies in recognition of and dealing with divided knowledge. It is reasonable that children know “what to do” if they are asked in the context of different school subjects. Thus, they are also aware of differences in the suitability of different sources of

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

5.00

Floating and sinking Sorting magnitudes Assign values

Sachkunde lesson Math lesson

Figure 6.1: Means of external sources ratings subdivided into curricular topics (significant differences between Sachkunde and math lesson context are marked)

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 181

knowledge for different tasks embedded in different school subjects. It is necessary to help students to disentangle the logic of a certain task from the subject matter context it is embedded in. Both aspects have to be considered when it comes to the evaluation of sources, but finding such a balance between both aspects has to be taught and practiced in order to develop an understanding of knowledge structures and their relationship to sources of knowledge.

It becomes obvious that students do not only have an idea of the dif­ference of knowledge within different disciplines and between different tasks; they also adjust their choice of sources to the respective discipline and task. However, it is still open if students would also know what to do when the context of different problems is not given. As outlined previ­ously, it is sometimes not so easy to find the most suitable discipline for a specific question. In this study, we examined students’ source prefer­ences by definition of two disciplines (math and Sachkunde lesson). With regard to implications for learning and instruction (see below), it is important to take into account that not all everyday problems are as well­framed and contextualized as they are in most typical school lessons.

Interestingly, in this study internal sources were preferred to exter­nal sources, which mirrors the prevailing assumption in most research

Book

5.0

4.5

3.5

2.5

1.5

0.5

4.0

3.0

2.0

1.0

0Friend

**

* *

*

Internet Parents

Sachkunde lesson Math lesson

Teacher Ownplanning

Try out Calculation

Figure 6.2: Mean source preferences for math and Sachkunde lesson context (significant differences between Sachkunde and math lesson context are marked)

Who knows what and who can we believe? 182

that knowledge constructed by the knower is often believed to be better than knowledge from outside. This is important in regard to the fact that much of our knowledge is acquired from others. We have argued beforehand that it is crucial in a knowledge society to depend on other sources, and the study provides some evidence that we should foster a more positive attitude toward this second­hand evaluation.

Implications for learning and instruction

The main implication for learning and instruction is the adoption of a certain perspective on learning and epistemological beliefs: as far as students are supported in understanding distributed knowledge struc­tures and, thereby, supported in their ability to assess the relevance and the veracity of knowledge claims, their epistemological beliefs are improved. We suggest another perspective on teaching epistemological beliefs that emphasizes that such beliefs do not primarily refer to the students’ own knowledge, but to the knowledge which is “offered” in the student’s environment (by textbooks, the Internet, peers, parents, television) in order to be attained by the student.

How could such beliefs about the structures of expert knowledge and, thereby, an understanding of distributed knowledge structures be improved? First of all, it is necessary to deliberately teach about the division of cognitive labor.5 As we have reviewed previously, such teach­ing does not need to start from scratch. It can be based on the basic understanding of the division of cognitive labor which has been scru­tinized by developmental psychologists. If teachers explicitly take into account major aspects of a knowledge society in their lessons, students’ attitudes toward second­hand evaluation could be improved. School education could lessen the utopistic burden of actively learning (and knowing) every piece of knowledge “essential” to being a responsible member of a knowledge society. As we have outlined before, so far, rely­ing on others in the sense of second­hand evaluations is not conceived of as adequate, as the normative goal most often is to overcome the division of cognitive labor. We would argue that as long as people judge their own relative knowledge and other people’s relative knowledge, the trustworthiness and relevance of sources, and relying on others, can be a well­adapted behavior, as most judgments about knowledge

5 For example, Sandoval (2005) enlisted epistemological themes that students should know in order “to be able to evaluate scientific claims in relation to socioscientific issues in their lives outside of (and beyond) school … Related to this is the notion that scientific knowledge is socially constructed, and thus includes cooperation, collabora­tion, and competition” (p. 639).

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 183

claims cannot be made by referring to one’s own personal experience. Students should be taught how to assess the relevance and the veracity of knowledge claims. A focus in learning and instruction on different genres of information transmission (texts) and types of sources could be helpful. One example includes different programs that have been invented explicitly to teach how to evaluate websites (e.g., www.2Learn.ca; Varnhagen, 2002).

Furthermore, developing epistemological beliefs for first­hand as well as second­hand evaluation of experts’ knowledge claims could be fostered by teaching critical reading. As most knowledge claims come along as texts (for example, on the Internet), “critical thinking” (an educational goal closely related to sophisticated epistemological beliefs) should be taught as “critical reading.” Such an emphasis on reading and comparing texts is typical for history (as a subject mat­ter discipline, but in history as a school subject not taught enough, according to Wineburg, 1991). However, it is not common in subject matter disciplines where texts are conceived of as less central than in the humanities.

Analyzing how knowledge claims are put into texts could be reveal­ing as well. For example, Smyth (2004) has demonstrated differing epistemological profiles of psychology textbooks in comparison to sci­ence textbooks. She states that science textbooks ususally present much more autonomous facts: statements are presented in present tense with­out any references or origins. In contrast, psychology textbooks anchor statements to evidence and support them by references. Therefore the focus in psychology textbooks is more on “what has been shown to be” rather than on “what is,” which implies that in psychology textbooks psychological entities are not treated as autonomous facts (Smyth, 2004, p. 530). One can conclude from Smyth’s work that the procedural and constructive character of scientific research is often neglected or at least underrepresented in science textbooks.

In research comparing students’ epistemological beliefs about psy­chology and science it can be concluded that such differences between ways of making knowledge claims really do matter. For example, Hofer (2000) compared students’ beliefs about knowledge in science with stu­dents’ beliefs about knowledge in psychology. Results show disciplinary differences, revealing that participants saw knowledge in science to be more stable, certain, and handed down by authority than in psychology. Furthermore, participants thought that truth in science is attainable by experts in contrast to the rather low attainability of truth in psychology. Hofer (2000) points out that the differences found between the differ­ent disciplines do not necessarily imply that students hold less advanced

Who knows what and who can we believe? 184

beliefs about science than about psychology, but instead might mirror views of the disciplines that professionals, would also likely share. Such disciplinary differences might also be caused by the way knowledge is treated in different subjects in schools or universities as well as in textbooks.

We have pointed out before that school subjects usually have to some degree a life of their own (see note 3), and this is also mirrored in the different ways knowledge is treated in lessons and textbooks. So far, the introduction to scientific modes of thinking and working plays a subordinate role in the teaching of natural sciences (at least in Ger­many; Prenzel and Parchmann, 2003), and the emphasis usually lies on imparting conceptual knowledge that teachers, authors of textbooks, and curricula take as established (Seidel et al., 2006). Only exception­ally do natural science lessons at school teach scientific controversies or the daily routine of research characterized by contradictory and fragile evidence (Labudde, 2000).

There is some evidence that considering argumentation and knowl­edge development in a specific discipline or knowledge domain might help to develop more adequate epistemological beliefs. Several researchers (Bell and Linn, 2002; Kuhn, 1991; Mason and Boscolo, 2004) stress the role of introducing controversies into science classes in order to promote students’ epistemological beliefs. Mason and Boscolo (2004) investigated the influence of high school students’ epistemological understanding on the critical interpretation of a dual­position text. Beforehand, participants’ epistemological under­standing was assessed using the instrument of Kuhn et al. (2000). Overall values were used to build three groups of different episte­mological positions, indicating whether participants primarily hold a less advanced, moderate, or more advanced view. All students read a scientific text about genetically modified food, introducing both the position in favor of and against this kind of food. After reading the text, participants were asked to write a conclusion to the text. Findings revealed that both students with more advanced beliefs and students with moderate epistemological understanding reflected bet­ter on the inconclusive nature of the debate on transgenic food. For example, they pointed out that more scientific studies on the topic are needed. Education should strive to produce sophisticated students, not only because of studies indicating a relation between advanced beliefs and dealing with controversies, but also because of a “univer­sal” need for a sophisticated standpoint in a knowledge­based society, as we have pointed out throughout this chapter (see also Bromme and Kienhues, 2008).

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 185

In a study on knowledge about genetics (Kienhues et al., 2008), we investigated the potential for influencing epistemological beliefs about genetics through a short instructional intervention, inspired by inter­vention strategies developed in the context of research on conceptual change. Based on an initial survey, two groups of university students were selected, one with less advanced epistemologies and the other with more advanced beliefs. A test of prior knowledge showed that both groups had little knowledge on the topic of genetics. Fifty­eight uni­versity students were randomly assigned to different conditions: one group whose epistemological beliefs were challenged through a refuta­tional epistemological instruction, which focused on the uncertainties and difficulties in DNA­fingerprinting. Another group which received a non­challenging informatory instruction outlining facts on DNA fin­gerprinting. The refutational epistemological instruction exemplifies that knowledge about genetics is controversial and uncertain or devel­oping. Because this instruction focuses on the questionable certainty of scientific propositions, it furthermore should provide a way to pro­voke epistemic doubt (Bendixen and Rule, 2004) in participants with less advanced beliefs. The treatment effect was assessed by comparing pre­instructional and post­instructional measures, using the CAEB (Stahl and Bromme, 2007), an instrument that convincingly meas­ures evaluative aspects of domain­specific epistemological beliefs. As hypothesized, the less advanced group receiving the refutational episte­mological instruction changed toward a more desirable advanced view. These participants’ former beliefs had been challenged through the instruction, which exemplified a more advanced epistemological view­point. In contrast, the more advanced group that received the informa­tory instruction changed toward a more naïve view. We were able to replicate comparable findings in two other studies.

The studies outlined previously provide empirical evidence for the need for more advanced epistemological beliefs in dealing with com­plex and perhaps conflicting information. This strengthens our view that being able to benefit from the division of cognitive labor includes epistemological beliefs. Furthermore, our study on genetics provides some first ideas about how epistemological beliefs can be improved. For example, it might be worthwhile to explicitly illustrate the inad­equacies and controversies of themes and theories to students and it might be advisable to avoid exaggerated claims regarding the certainty of knowledge. This view is confirmed by the finding that the sophis­ticated group receiving the uncontroversial informational instruction shifted significantly toward a more naïve view, and it is also in line with the assumptions on textbooks that we outlined earlier.

Who knows what and who can we believe? 186

Different disciplines often differ remarkably with regard to their ontology and, therefore, in their ways of establishing “truth.” While “goals” and “intentions” could be conceived of as legitimate explana­tions for the behavior of humans, it would be a categorical error to refer to “intentions” as explanations for the behavior of physical entities such as stones, light waves, and molecules (Keil, 2006). Only when someone knows the questions a discipline deals with is she/he able to know the features which have to be taken into account for the assessment of cer­tainty (Bromme et al., 2008).

A starting point for learning about the questions a specific discipline deals with could be to foster a comparative approach on different school subjects. For example, Stevens et al. (2005) have argued for a compara­tive approach to study subject matter learning in school. They propose to consider how students’ experiences, embedded in temporal, material, emotional, and social organizational conditions of the school day, frame students’ learning of academic subjects. Taking these conditions into account, subject matter teachers could support students to learn what is unique about a discipline and how it could be compared with others.

To conclude, different aspects should be taken into account in learning and instruction: an understanding of distributed knowledge structures should be as well supported as the ability to assess the rel­evance and the veracity of knowledge claims. Different aspects might add to these competences, such as critical reading, an adequate con­sideration of the potential preliminarity and changeability of knowl­edge in textbooks, advanced epistemological beliefs, and ontological knowledge.

Implications for further research

Many current conceptualizations of personal epistemological beliefs do not keep apart epistemological beliefs about the subjects’ own knowl­edge and their beliefs about the knowledge of others. “Personal” is an ambiguous term as it can refer to the person as the subject (who holds the beliefs about knowledge), and “personal” can also refer to the object of such beliefs (the holder of the knowledge itself). A student might have subjective epistemological beliefs about: (a) how scientists (the objects of her/his beliefs) produce and justify knowledge about science topics, and she or he might have beliefs about (b) their own ways of acquiring and justifying knowledge about science topics.

The widely used definition of personal epistemology as four dimen­sions of beliefs about knowledge and knowing is neutral against the questions whose knowledge such beliefs refer to. Questioning this

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 187

neutrality, some authors have emphasized the need for being more spe­cific with regard to whose knowledge personal epistemological beliefs refer to. Hogan (2000) has suggested the notion of distal epistemologi­cal knowledge; Sandoval (2005) has suggested the notion of formal epistemology in order to characterize the first. The notion of proximal (Hogan, 2000) and practical epistemology (Sandoval, 2005) has also been proposed in order to describe the second understanding of per­sonal epistemological beliefs.

The conceptual distinction suggested by these authors is very helpful in order to clarify the main message of this chapter: while most recent research on epistemological beliefs either remains ambiguous with regard to the personal character of epistemological beliefs, or focuses on proximal beliefs or practical epistemology by asking for students’ ideas about their own knowledge, we argue for clearer distinctions. We suggest future research on subjects’ beliefs about distal knowledge, especially about the division of cognitive labor. As outlined throughout this chapter, in a more and more complex world second­hand evalua­tions are most common, as first­hand evaluations are a rare exception. Distal knowledge (or formal epistemology) is crucial to equip individu­als in doing this type of evaluation.

Future research should not only help to build a clearer picture of people’s distal knowledge; it should also examine ways to foster it. For example, Hogan (2000) suggests that students’ distal knowledge could be instilled by urging them to reflect on the history of science, like the role of controversies or the role of society and culture in verifying knowledge claims. Whether learning about the philosophy and history of science is indeed helpful is still an open empirical question. It is also an open question in how far such distal knowledge should be fostered in a more general or highly discipline­specific way.

The studies outlined before indicate that children already have a basic understanding that knowledge in the minds of others is clustered in a discipline­based way, but it is open in how they would show a critical reflection of knowledge claims. Knowing who knows what and which source is suitable for what is only half the battle, as we have pointed out. For example, information on the web is unregulated, therefore high­quality web sites are not always easy to find, and just knowing that there might be suitable information “somewhere on the web” is not enough to find a specific answer. Although there is some research indicating that children have a basic ability to identify incorrect infor­mation and to judge the trustworthiness of information, there are also various studies indicating that people are not always capable or will­ing to make adequate source judgments. Investigating under which

Who knows what and who can we believe? 188

conditions people take the trustworthiness or validity of information for granted would be a good starting point for interventions to reduce such blind trust in others.

It might also be fruitful for future research to take into account theories of document representation (e.g., Perfetti et al., 1999). The increased availability of fragile and contradictory evidence in a knowl­edge society often requires the integration of information from multiple documents (Stadtler and Bromme, 2008). In general, multi­layered, intertwined approaches, for example taking into account document representation, epistemological beliefs, and an understanding of the division of cognitive labor, will hopefully offer more comprehensive insights.

Conclusion

We do not deny that researching students’ ideas about their own knowledge is relevant for improving their epistemological beliefs, but we think that the perspective adopted here could help to maintain a clear focus on the place of epistemological beliefs within the capaci­ties which are necessary for lifelong learning from others and with the lifelong need to deal with specialized expert knowledge without being an expert oneself. Supporting students’ understanding of knowledge structures within their society (and also within other societies and cultures, Khine, 2008), fostering a comparative view on school sub­jects, teaching critical reading of the typical and discipline­specific text genres of knowledge claims, and also teaching the assessment of knowledge sources all focus on beliefs about knowledge provided by and learned from others. Doing this successfully will also, of course, improve the epistemological beliefs about the students’ own knowledge.

Acknowledgments

We would like to thank Lisa Bendixen and Florian Feucht for their helpful comments on an earlier version of the chapter.

R EF ER ENCES

Baxter Magolda, M. B. (1992). Knowing and reasoning in college: Gender-related patterns in students’ intellectual development. San Francisco: Jossey­Bass.

Bergstrom, B., Moehlmann, B., and Boyer, P. (2006). Extending the testimony problem: Evaluating the truth, scope, and source of cultural information. Child Development, 77, 531–8.

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 189

Bell, P. and Linn, M. C. (2002). Beliefs about science: How does science instruction contribute? In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (321–46). Mahwah, NJ: Lawrence Erlbaum Associates.

Bendixen, L. D. and Rule, D. C. (2004). An integrative approach to personal epistemology: A guiding model. Educational Psychologist, 39, 69–80.

Buehl, M. M. and Alexander, P. A. (2006). Examining the dual nature of epistemological beliefs. International Journal of Educational Research, 45, 28–42.

Bransford, J. D., Brown, A. L., and Cocking, R. R. (Eds.) (2000). How people learn: Brain, mind, experience, and school. Washington, DC: National Acad­emy Press.

Bråten, I., Strømsø, H. I., and Samuelstuen, M. S. (2005). The relationship between internet­specific epistemological beliefs and learning within internet technologies. Journal of Educational Computing Research, 33, 141–71.

Bromme, R. (1994). Beyond subject matter: A psychological topology of teach­ers’ professional knowledge. In R. Biehler et al. (Eds.), Mathematics didactics as a scientific discipline: The state of the art (73–88). Dordrecht: Kluwer.

Bromme, R., Jucks, R., and Runde, A. (2005). Barriers and biases in computer­mediated expert­layperson­communication. In R. Bromme et al. (Eds.), Barriers, biases and opportunities of communication and cooperation with com-puters – and how they may be overcome (89–118). New York: Springer.

Bromme, R. and Kienhues, D. (2008). Allgemeinbildung [general educa­tion]. In W. Schneider and M. Hasselhorn (Eds.), Handbuch der Päda-gogischen Psychologie [Handbook of Educational Psychology] (619–28). Göttingen: Hogrefe.

Bromme, R., Kienhues, D., and Stahl, E. (2008). Knowledge and epistemo­logical beliefs: An intimate but complicate relationship. In M. S. Khine (Ed.), Knowing, knowledge and beliefs (423–41). New York: Springer.

Brownlee, J. (2003). Changes in primary school teachers’ beliefs about know­ing: A longitudinal study. Asia-Pacific Journal of Teacher Education, 31, 87–98.

Danovitch, J. H. and Keil, F. C. (2004). Should I ask a fisherman or a biologist?: Developmental shifts in ways of clustering knowledge. Child Development, 75, 918–31.

Donald, J. G. (1995). Disciplinary differences in knowledge validation. In N. Hativa and M. Marincovich (Eds.), Disciplinary differences in teaching and learning: Implications for practice (7–17). San Francisco: Jossey­Bass.

Eysenbach, G. and Köhler, C. (2002). How do consumers search for and appraise health information on the world wide web? Qualitative study using focusing groups, usability tests, and in­depth interviews. British Medical Journal, 324, 573–7.

Goswami, U. (2008). Cognitive development: The learning brain. Hove: Psychol­ogy Press.

Haerle, F. C. and Bendixen, L. D. (2008). Personal epistemology in elementary classrooms: A conceptual comparison of Germany and the United States

Who knows what and who can we believe? 190

and a guide for future cross­cultural research. In M. S. Khine (Ed.), Knowing, knowledge and beliefs (151–76). New York: Springer.

Hammer, D. and Elby, A. (2003). Tapping epistemological resources for learn­ing physics. The Journal of the Learning Sciences, 12, 53–90.

Harris, P. L. (2001). Thinking about the unknown. Trends in Cognitive Sciences, 5, 494–8.

Hofer, B. K. (2000). On dimensionality and disciplinary differences in per­sonal epistemology. Contemporary Educational Psychology, 25, 378–405.

(2008). Personal epistemology and culture. In M. S. Khine (Ed.), Knowing, knowledge and beliefs: Epistemological studies across diverse cultures (3–22). New York: Springer.

Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learn­ing. Review of Educational Research, 67, 88–140.

Hogan, K. (2000). Exploring a process view of students’ knowledge about the nature of science. Science Education, 84, 51–70.

Irwin, A. and Wayne, B. (Eds.) (1996). Misunderstanding science? The pub­lic reconstruction of science and technology. Cambridge: Cambridge University Press.

Kalish, C. W. (2002). Essentialist to some degree: Beliefs about the structure of natural kind categories. Memory and Cognition, 30(3), 340–52.

Keil, F. C. (2006). Explanation and understanding. Annual Review of Psychol-ogy, 57, 227–54.

Keil, F. C., Stein, C., Webb, L., Billings, V. D., and Rozenbilt, L. (2008). Discerning the division of cognitive labor: An emerging understand­ing of how knowledge is clustered in other minds. Cognitive Science, 32, 259–300.

Khine, M. S. (Ed.) (2008). Knowing, knowledge and beliefs: Epistemological stud-ies across diverse cultures. New York: Springer.

Kienhues, D., Bromme, R., and Stahl, E. (2008). Changing epistemological beliefs: The unexpected impact of a short­term intervention. British Jour-nal of Educational Psychology.

King, P. M. and Kitchener, K. S. (1994). Developing reflective judgment: Under-standing and promoting intellectual growth and critical thinking in adolescents and adults. San Francisco: Jossey­Bass.

Kuhn, D. (1991). The skills of argument. Cambridge: Cambridge University Press.

Kuhn, D., Cheney, R., and Weinstock, M. (2000). The development of episte­mological understanding. Cognitive Development, 15, 309–28.

Kuhn, D. and Weinstock, M. (2002). What matters in epistemological thinking and why does it matter? In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (121–44). Mahwah, NJ: Lawrence Erlbaum Associates.

Labudde, P. (2000). Konstruktivismus im Physikunterricht der Sekundarstufe II [Constructivism in physics lessons at the high school level]. Bern, Stutt­gart, Wien: Verlag Paul Haupt.

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 191

Limon, M. (2006). The domain generality­specificity of epistemological beliefs: A theoretical problem, a methodological problem or both? Interna-tional Journal of Educational Research, 45, 7–27.

Liu, D., Wellman, H. M., Tardof, T., and Sabbagh, M. A. (2008). Theory of mind development in Chinese children: A meta­analysis of false­belief understanding across cultures and languages. Developmental Psychology, 44, 523–31.

Lutz, D. J. and Keil, F. C. (2002). Early understanding of the division of cogni­tive labor. Child Development, 73(4), 1073–84.

Mason, L. and Boldrin, A. (2008). Epistemic metacognition in the context of information searching on the web. In M. S. Khine (Ed.), Knowing, knowl-edge and beliefs: Epistemological studies acrosse diverse cultures (377–404). New York: Springer.

Mason, L. and Boscolo, P. (2004). Role of epistemological understanding and interest in interpreting a controversy and in topic­specific belief change. Contemporary Educational Psychology, 29, 103–28.

Muis, K. R. (2004). Personal epistemology and mathematics: A critical review and synthesis of research. Review of Educational Research, 47(3), 317–77.

Olafson, L. and Schraw, G. (2006). Teachers’ beliefs and practices within and across domains. International Journal of Educational Research, 45, 71–84.

Perfetti, C. A., Rouet, J.­F., and Britt, M. A. (1999). Toward a theory of doc­uments representation. In H. v. Oostendorp and S. R. Goldman (Eds.), The construction of mental representations during reading (99–122). Mahwah, NJ: Lawrence Erlbaum Associates.

Perry, W. G. (1970). Forms of intellectual and ethical development in the college years: A scheme. San Francisco: Jossey­Bass.

Porsch, T., Bromme, R., and Pollmeier, J. (2010). Was muss man tun um sicher die richtige Lösung zu finden? Quellenbeurteilungen von Grundschulkindern in verschiedenen Fachkontexten [What to do in order to find the correct solution? Source judgements of primary school students]. Zeitschrift für Entwicklungspsychologie und Padagogische Psychologie [Journal for Develop-mental and Educational Psychology].

Prenzel, M. and Parchmann, I. (2003). Kompetenz entwickeln: Vom naturwis­senschaftlichen Arbeiten zum naturwissenschaftlichen Denken: Natur­wissenschaften im Unterricht: [Developing competency: from scientific practicing to scientific thinking: Sciences in education]. Chemie [Chemis-try], 14, 169–71.

Rozendaal, J. S., de Brabander, C. J., and Minnaert, A. (2001). Boundaries and dimensionality of epistemological beliefs. Paper presented at the 9th Euro­pean Conference for Research on Learning and Instruction, Fribourg, Switzerland.

Runde, A., Bromme, R., and Jucks, R. (2007). Scripting in net­based medi­cal consultation: The impact of external representations on giving advice and explanations. In F. Fischer et al. (Eds.), Scripting computer-supported collaborative learning: cognitive, computational, and educational perspectives (57–72). New York: Springer.

Who knows what and who can we believe? 192

Sandoval, W. A. (2005). Understanding students’ practical epistemologies and their influence on learning through inquiry. Science Education, 89, 634–56.

Schwab, J. J. (1987). Education and the structure of disciplines. In I. Westbury and N. J. Wilkof (Eds.), Science, curriculum, and liberal education (229–272). Chicago: University of Chicago Press.

Seidel, T., Prenzel, M., Rimmele, R., et al. (2006). Blicke auf den Physikun­terricht: Ergebnisse der IPN Videostudie [Views on physics educa­tion: Results from the IPN video study]. Zeitschrift für Pädagogik [Journal of Pedagogy], 52(6), 799–821.

Smyth, M. M. (2004). Exploring psychology’s low epistemological profiles in psychology textbooks. Theory and Psychology, 14, 527–53.

Stadtler, M. and Bromme, R. (2007). Dealing with multiple documents on the www: The role of metacognition in the formation of documents mod­els. International Journal of Computer Supported Collaborative Learning, 2, 191–210.

(2008). Effects of the metacognitive tool met.a.ware on the web search of laypersons. Computers in Human Behavior, 24, 716–37.

Stahl, E. and Bromme, R. (2007). The CAEB: An instrument for measuring connotative aspects of epistemological beliefs. Learning and Instruction, 17, 773–85.

Stehr, N. (1994). Knowledge societies. London: Sage.Stevens, R., Wineburg, S., Herrenkohl, L. R., and Bell, P. (2005). Compara­

tive understanding of school subjects: past, present, and future. Review of Educational Research, 75, 125–57.

Stodolsky, S. S. (1988). The subject matters: Classroom activity in math and social studies. Chicago, IL: University of Chicago Press.

Strømsø, H. I., Bråten, I., and Britt, M. A. (2008). Do students’ beliefs about knowledge and knowing predict their judgment of texts’ trustworthiness? Manuscript submitted for publication.

Tsai, C.­C. and Chuang, S.­C. (2005). The correlation between epistemologi­cal beliefs and preferences toward internet­based learning environments. British Journal of Educational Technology, 36, 97–100.

Tuckett, D. A., Boulton, M., Olson, C., and Williams, A. (1985). A new approach to the measurement of patients’ understanding of what they are told in medical consultations. Journal of Health and Social Behavior, 26, 27–38.

Varnhagen, C. K. (2002). Making sense of psychology on the web: A guide for research and critical thinking. New York: Worth Publishers.

Wellman, H. M., Cross, D., and Watson, J. K. (2001). Meta­analysis of theory­of­mind development: The truth about false belief. Child Development, 72, 655–84.

Whitmire, E. (2004). The relationship between undergraduates’ epistemologi­cal beliefs, reflective judgment, and their information­seeking behavior. Information Processing and Management, 40, 97–111.

Wimmer, H. and Perner, J. (1983). Beliefs about beliefs: Representation and constraining function of false beliefs in young children’s understanding of deception. Cognition, 13, 103–28.

Rainer Bromme, Dorothe Kienhues, and Torsten Porsch 193

Wineburg, S. S. (1991). Historical problem solving: A study of the cognitive processes used in the evaluation of documentary and pictorial evidence. Journal of Educational Psychology, 83, 73–87.

Wu, Y.­T. and Tsai, C.­C. (2007). Developing an information commitment survey for assessing students’ web information searching strategies and evaluative standards for web materials. Educational Technology and Society, 10, 120–32.

Part III

Students’ personal epistemology, its development, and its relation to learning

197

7 Stalking young persons’ changing beliefs about belief

Michael J. Chandler and Travis ProulxThe University of British Columbia

Introduction

Somewhere very near the top of psychology’s list of most vexing and least settled matters are the nagging questions of how and when young persons come to anything like a “mature” account (or folk conception) of the nature and limitations of human knowing. That is, we (where “we” refers to all of us salaried professionals actually paid to know about such things) seem unable to agree about almost anything hav­ing to do with people’s changing beliefs about belief (Chandler et al., 2002; Chandler and Sokol, 1999). Do young persons ordinarily aban­don some entry­level commitment to “naive realism” at the age of four, or is it fourteen, or twenty­four (Chandler and Carpendale, 1998)? Are our earliest insights about the inherently agentic (and therefore ine­luctably subjectivized or relativized) nature of human knowing stand­ardly acquired during the preschool or, rather, the post­graduate years (Chandler et al., 2000)? When, give or take a few decades, is it fair to say that young persons will have already acquired a journeyman’s “theory­ of­mind,” adult­like in all of its basic particulars (Chandler, 2001)? Is it four, or eight, or twelve, and, if not, do such accomplishments await some “age of majority,” or the acquisition of a liberal arts degree (Kitchener and King, 1981)? No fair­minded reader of the contempo­rary research literature on personal epistemologies could, we maintain, come away from an exhaustive review of the several hundred studies given over to such matters with anything like a confident conclusion (Chandler et al., 2002).

Further purple prose aside, something clearly needs to be done here. Our collective conceptual house is in evident disarray, and in desperate need of being put in some better order. This chapter is intended, not as some premature way of actually solving this problem once and for all – obviously a task too big for the evidence available and the space

Stalking young persons’ changing beliefs198

allotted – but is meant, instead, to simply provide certain conceptual tools that will hopefully prove useful in getting this house cleaning job underway.

In search of a reasonable place to start this tool­gathering enter­prise, we propose beginning with some lining­out of those relatively few matters on which professional and lay opinion would seem to con­verge. For example, no one appears to seriously doubt that toddlers and ruddy­faced preschoolers seem generally predisposed to backing their entry­level knowledge claims by simply pointing to what they take to be self­evident facts in the world. That is, given their apparent doctrine of “immaculate perception” (Chandler and Lalonde, 1996), what seem­ingly strikes these novice epistemologists as the right way to back some presumed fact (e.g., that “cherries are red” or that “balls are round”) is to simply direct your attention to what they evidently take to be ines­capable perceptual proof­positive. By contrast, and on the far side of any such fledgling insights into the knowing process, is the hoped for prospect that, given world enough and time, more fully fledged ado­lescents and young adults will ordinarily come to some much headier, context­sensitive appreciation of the intricacies of what knowing could possibly be all about. It is, no doubt, in the spirit of such “older is better” presumptions that we commonly invent laws about appropriate “ages of majority,” and insist that judges and other candidates for high office not be entirely wet behind the ears. In short, we standardly make pro­vision for the taken­for­true prospects that anyone still in short pants will likely prove to be naïve (perhaps even naïvely realist), while antici­pating that your average grown­up will ordinarily read more relativized subtlety into their increasingly more “mature” folk epistemologies than do their offspring. Beyond these extremes, however, all clear consensus would seem to break down – not just among the general citizenry, but also among psychology’s leading contributors to the literature on so­called “personal epistemologies” (Chandler et al., 2002).

On reducing the range of possible alternatives

At one extreme in this array of seemingly endless possibilities, whole armies of contemporary “theory theorists” insist, for example, that, by four or five, the typical preschooler has already acquired a “theory of mind” that is, in all important respects, a finished product, not dif­ferent in kind from those later­arriving “theories” practiced by most adults (Perner, 1991). For other investigators, by contrast – typically those whose research “subjects” tend to be ready­to­hand college stu­dents and other candidates for higher degrees – anything approaching a

Michael J. Chandler and Travis Proulx 199

mature grasp of epistemic matters is widely thought to be the exclusive province of “upper­class men” [sic], or even those already gifted with some higher degree (Hofer and Pintrich, 1997; King and Kitchener, 2002; Perry, 1970). To the “degree” that this later­is­better view (the one linking epistemic sophistication to post­secondary education) is seen to gain research support, the stock of “higher” education, and those who provide it, necessarily goes up.

Perhaps, some of all of this diversity of professional opinion is actu­ally a sign of health. After all, science is said to thrive on disagreement. In this spirit, no one would likely be surprised or disappointed to learn, for example, that this or that key epistemic ability was differently said to put in its earliest appearance at three, as opposed to five, or fifteen, as opposed to twenty. These are not, however, the near­miss differences of professional opinion that must trouble us here (Chandler et al., 2002). Rather, as things currently stand, it is hard to countenance competing claims about competing knowledge claims that regularly differ, as is standardly the case in the available epistemic literature, by as much as two or more decades. Either some of our research colleagues are con­fusedly calling radically different things by the same name, or someone has obviously gotten their facts badly wrong. Clearly, steps for deciding among these unhappy alternatives are sorely needed. One way to begin guiding such steps is to try for enough conceptual altitude to get the general lie of this disputed landscape.

A high altitude bombing run

Reconnoitering from a viewing distance sufficiently rarified to poten­tially allow for the prospect of envisioning a developmental arc large enough to potentially encompass the whole of childhood, it is possible, we and occasionally others (e.g., Chandler et al., 2002; Hallett et al., 2002; Kuhn and Weinstock, 2002) have argued, to identify two dis­tinct, but not quite orthogonal, dimensions along which most, or all, of the currently available knowledge claims about competing knowledge claims can be seen to fall.

One of these possibilities stakes out a continuum of expectations concerning the anticipated likelihood that any given assertion of fact will garner high, as opposed to low, levels of inter-subjective agreement. Fifty million Frenchmen, we are regularly cautioned, are probably not wrong. Consequently, we might, and others have (Chandler et al., 2002; Hallett et al., 2002; Kuhn and Weinstock, 2002; Pillow, 1991, 1999; Schommer, 1990), anticipated that an important dimension cross­cut­ting any age­graded variations in what young persons think about the

Stalking young persons’ changing beliefs200

nature and course of human knowledge will somehow turn upon some notion of epistemic consensus.

The second of these proposed dimensions runs deep rather than long, moving from something like “internality” to something like “external­ity” (Chandler et al., 2002; Hallett et al., 2002; Kuhn and Weinstock, 2002; Pillow, 1999) – a dimension meant to carve out a space within which people are expected to actually point when locating the authority required to back this or that particular knowledge claim. At one pole of this second, spatialized dimension it is possible to imagine some free­standing, environing world – a world thought capable of dictating the non­negotiable terms of any and all of one’s “primary” sense experi­ences – what Kuhn and Weinstock (2002) call “physical truths.” In the extreme, such self­evident facts are often thought to belong to a world of “brute facts,” common to all with eyes to see (Rorty, 1991). At the other and more “internalized” limit, one’s accusing finger is thought to curl back upon itself, pointing to some interiorized and often highly private place where “gut level” knowledge, and intuitive or instinctive self­understandings, are thought to reside, lending legitimacy to one’s personal beliefs, values and preferences (Searle, 1983).

Reduced to such bare­bones, lower­bound limits, the crossing of these two imagined epistemic dimensions yields a classic two­by­two contingency table (see Table 7.1) in which only one of the available diagonals is imagined to be occupied.

That is – within the architecture of this entity­level folk epistemol­ogy – there are imagined, on the one hand, to be sensory–based, more or less indisputable, perceptual knowledge claims (e.g., this is red, that is blue) that reputedly draw their presumably uncontestable authority from the widely shared “brute facts” (Searle, 1983) that literally make up the environing world. On the other hand, and in the remaining cell of this diagonal, there is routinely thought to be a residual place for housing all of those often equally heart­felt beliefs (e.g., chocolate tastes better than vanilla) that draw their supposed authority from some more private, internalized place – an idiosyncratic circumstance that robs them of any real hopes for necessary inter­subjective agreement, and

Table 7.1. Two imagined epistemic dimensions

High consensus Low consensus

Internal justification Social facts Tastes

External justification Brute facts

Michael J. Chandler and Travis Proulx 201

(some would claim) any of the legitimately grounded epistemic status required for being bona fide “facts” of any sort (Elgin, 1989).

If something like the above was all there were to it – that is, if the known world really did conveniently divide itself into easily settle­able matters of brute facts on the one hand, and un­arbitrateable matters of personal taste on the other, then it would follow that there is an auto­matic and seriously simplifying one­to­one correspondence between (sometimes still pending) inter­subjective agreement and objectivity. Relying on any such minimally complex personal epistemology, we would, for example, simply overrule those who claim to see matters dif­ferently by confronting them with still more brute facts, while showing a laudable tolerance for all of those fully anticipated, but completely incorrigible, differences of opinion, presumably owed to the fickle and highly personalized nature of aesthetic preferences.

Among the potential reasons that something like this entry­level, but still often serviceable, either/or division of labor between matters of fact and matters of taste has proved to be so attractive is that, were it true, all that would be required, in order to be a competent and epistemologi­cally mature practitioner of this trade, would be a set of cognitive abili­ties that are generally understood to be well within the ability range of your average three­ or four­year­old – children who already have simple facts down pat, and who already seem content with the idea that cats really do like smelly cat food (Flavell et al., 1992), or that it is fine to offer hated things like broccoli to adults who evidently have a taste for them (Repacholi and Gopnik, 1997).

A rather long list of reasons militate against anything quite as simple as any such short­pants folk­epistemology – one that leaves no room for anything beyond a world of consensually agreed upon brute facticity, and a divided world of aesthetic opinions. Not the least of these doubts is our largely unshakeable common conviction that there must be more to the course of epistemic development than preschoolers ordinarily have eyes to see.

One of these concerns is that our human world is generally under­stood to be full to overflowing not just with unadjudicated opinions and so­called “brute” facts – facts that supposedly defy reasonable disagreement – but also includes both shared “values” (Elgin, 1989) or “social truths” (Kuhn and Weinstock, 2002) and what John Searle (1983) has termed “social” or “institutional” facts. Facts such as “the Giants beat the Dodgers 3 to 2,” or that “Jon and Jane are married” – real facts­of­the­matter that could only be taken to be true in the con­text of some localized, historicized, humanly constructed rule system or socio­cultural frame of reference.

Stalking young persons’ changing beliefs202

Such shared values and social/institutional facts serve, we suggest, to help put the lie to any simple two­by­two explanatory structure that equates the possibility of consensus, on the one hand, with reliance on simple sense­based experience, on the other. This follows, in large part, from that fact that it is broadly constitutive of what values and social facts are standardly taken to be that they are, at the same time, both (a) inter­subjectively agreed upon and (b) humanly constructed. As such, they find no comfortable place in any simple two­by­two scheme that insists that the very possibility of real inter­subjective agreement requires some point­to­able referent in the material world.

Assuming that, for reasons of immaturity or inexperience, one con­tinues to be stuck with something like the sort of entity­level episte­mology outlined above, two general sorts of solution strategies appear available. One of these is to argue that what is being promoted as no more than a locally agreed upon, socially relativized value or insti­tutional “fact” must, on closer inspection, be revealed to be either (a) some disguised and previously unappreciated “brute” fact of nature – a sensory­motor fact ultimately available to anyone with the eyes to see it for what it actually is – or (b) alternatively, actually some matter of taste wrongly masquerading as a fact of nature. Something like the first of these moves was, for example, no doubt had in mind by Thomas Jefferson and other “founding fathers,” who promoted the democratic values underpinning their newly crafted US constitution as absolute “truths” – truths that he insisted were “self­evident.” Simi­larly, the proposition that a marriage is, and can only be, “a union between a man and a woman” is likewise meant to help clear the decks of other pesky interpretive possibilities that might reduce objective truths, supposedly sanctioned by God, to appear as no more than mere social conventions. In short, it is often imagined possible, at least in principle, to work to empty out any troublesome middle­ground cat­egory of publicly agreed upon social facts and values by branding some as underappreciated brute facts, while discounting any remainder as mere aesthetic preferences.

Militating for a middle way

The alternative – the one that we (along with others) mean to militate for in this chapter – is that, in addition to those all but universally agreed upon knowledge forms reputedly owed to more or less direct sense per­ception, and a residual set of often individualized claims regarding pri­vate truths that are said to be rooted in matters of personal taste, there also exists a further class of knowledge claims having to do with shared

Michael J. Chandler and Travis Proulx 203

values and social facts that, while traceable to no sort of exceptionless brute fact, nevertheless enjoy, as constitutive conditions of their exist­ence, a context­dependent form of at least local consensus. Much of the troubled character of epistemic development (both for those who are required to live through it, and those social scientists aiming to describe it), we (along with others) hope to show, is owed to the difficult task of finding ways of getting clear about the place of such humanly constructed, yet broadly agreed upon, social facts and values.

The beginnings of a plan

Given the foregoing account, we mean to go on to argue for two broad propositions. The first of these is that much of the confusion that con­tinues to plague both lay opinion and the research literature on epis­temic development grows out of a failure to properly acknowledge the existence and legitimate boundary considerations surrounding an inter­mediate category made up of various shared values and social facts – facts and values that set themselves apart from both easily agreed upon sense­based assumptions about the material world, on the one hand, and always inter­subjectively disputed claims about matters of aesthet­ics or tastes, on the other (Baxter Magolda, 1992; Kuhn and Wein­stock, 2002).

Second, we mean to argue the case that, while not entirely free of every developmental consideration, young children commonly succeed in hammering out some serviceable folk­epistemology capable of man­aging disputes about matters of brute facts and personal tastes at a very early age (generally in preschool and earliest school years). As such, there is no evident developmental arc, and so no compelling develop­mental story to be told about such matters. Rather, their (our) ongoing problems lie in working out what to do with contested matters of social facts and shared values – that odd­ball category of knowledge claims (e.g., “Jon and Jane are married”) that regularly enjoy some common consensus, but have no easily point­to­able external referent.

We mean to go about this re­framing project in three steps that will form the outline for the remainder of this chapter. Steps 1 and 2 will consist of efforts to selectively review empirical studies and interpretive commentaries that bear upon our contention that, from a surprisingly early age, young persons are already clear enough in their convictions that brute facts all but speak for themselves, and that what is inter­esting about knowledge concerning aesthetic matters rarely depends upon close inter­subjective agreement. Although we will work here to leave room for the prospects that general increases in age or cognitive

Stalking young persons’ changing beliefs204

complexity do apparently impact on what brute facts and aesthetic pref­erences are generally taken to be, our broad purpose will be to line up evidence in support of our competing argument that there is not a great deal of light between the ways that preschoolers and college graduates actually imagine themselves to understand both of these matters. Here our tasks will be to show: (a) that familiar claims from the literature that young persons gradually assemble some “evaluativist stance” toward matters of material facts may not, on closer inspection, prove to be what they are commonly advertised as being; and (b) that much of what is held up as being about issues of aesthetic preference are better seen as disputes about social facts and values. Once this redescription of avail­able evidence is accomplished, our plan will be to move on, in Step 3, to the final matter of shared values and social facts – the place where, at least in our scheme of things, most of the real business of epistemic development actually occurs.

Step 1: Matters of aesthetic judgment and personal taste

Before going too far down any road leading to discussions about pos­sible differences in whatever naïve epistemologies young people might harbor regarding disputed matters of taste, it seems important to first begin by acknowledging the existence of a widely held, but spoiler, view according to which any and all locutions aimed at bracketing together epistemic and aesthetic matters are held to be frankly oxymoronic, all for the reason that aesthetic claims are sometimes said to lack any epis­temic import whatsoever (i.e., they are not, on this more careful view, seen to have any truth­bearing relation to anything in the environing world). Of course, we commonly say things such as “I believe that the music of Beethoven is better than that of Brahms,” or that “It is com­mon knowledge that red wine is superior to white,” but, on the account currently under scrutiny, such turns of phrase are seen to be entirely self­referential and should not be imagined to make some defensible claim upon the truth.

However pristine one happens to find such impossibility arguments, for those concerned with folk­psychological matters, three things, at least, argue against any view that works to discount the very possibility that aesthetic claims possess real epistemic character. One of these is that there is no shortage of competing philosophical arguments all to the effect that aesthetic matters do, in point of fact, often make claims to truth. Hans­Gorge Gadamer’s highly influential Truth and method (1960/1982), for example, is only one of a whole genera of such accounts

Michael J. Chandler and Travis Proulx 205

that provide serious challenges to the empty abstractions inherent in many classic claims about aesthetic consciousness – insights meant to help redeem the possibility that matters of taste can and do regularly make warrantable claims to truth. Why, otherwise, are we interested in the views of theater or food critics? Such views have importantly impacted, not only on contemporary philosophy, but also theology, legal theory, and much of literary criticism. In short, before leaping on any philosophical high­horse in the hopes of riding rough­shod over presumptively misspoken claims about the very possibility of a folk­epistemology of aesthetics, it is worth pausing long enough to note that it is not possible to do so without fear of contradiction from other equally well­shod philosophical quarters (e.g., Gadamer, 1960/1982).

Second, and hopefully beyond any further “he said–she said” appeals to authority, it would also seem wise to note that, without stumbling into self­contradiction, one may not lightly dismiss the very possibility that aesthetic judgments are at least sometimes truth­bearing, and, at the same time, happily participate in the culture of connoisseurship that so permeates our complex social lives. The next time you double­ check with a theater or music critic before laying down your hard­earned cash, or find yourself turning to some fashion maven or wine expert before next choosing what you intend to wear or drink, take the time to rethink any lingering convictions you might entertain about the impossibility of there being some epistemic dimension to such mat­ters of taste. Presumably, such well­paid arbitrators of taste (including, of course, academics paid to offer up their personal “take” on this or that research literature) actually sometimes really do know something, and we often dismiss their advice at our own peril.

Finally, as the empirical work we intend to cite next makes clear, the folk­psychological accounts featured in this volume not only regularly operate as though aesthetic claims are sometimes either true or false, but go so far as to imagine that any claim that truth is really no better than a blind liking is, at worst, a simple affectation. With at least this much authority, then, it seems prudent to search out what little research evidence appears to exist about the developmental course of aesthetic judgments.

Aesthetic epistemologies at an early age

As previously hinted at in the Introduction, there already exists a small but impressive body of good evidence indicating that, well before school age, young children ordinarily appreciate that they and others often display different tastes and aesthetic preferences. More than twenty

Stalking young persons’ changing beliefs206

years ago, for example, Flavell et al. (1992) demonstrated that, by three or four years of age, preschoolers clearly appreciate that, while they personally find the prospect of eating cat food decidedly “yucky,” cats can be expected to have just the opposite reaction. Similarly, and more recently, Repacholi and Gopnik (1997) succeeded in pushing this threshold to still younger ages by showing that children as young as eighteen months are already well aware that it would be an act of kind­ness to offer broccoli to adults who have expressed a liking for it, as opposed to the goldfish crackers that they themselves prefer.

No parent is likely to be caught entirely off guard by such findings. That grown­ups often do put childish things behind them seems widely appreciated, even by young children, who appear to regularly expect that adults like and want adult things, while they, themselves, more often prefer the stuff of childhood. While some of these findings do, no doubt, lend themselves to more associationistic or reductive explana­tions – theories of “desire” rather than theories of “belief,” for example (Wellman, 1990) – all such claims not withstanding, it already seems clear enough that no one need wait until they have reached voting age, or matriculated into college, before first appreciating that differences in aesthetic preferences are the rule rather than the exception. Rather, from an especially tender age, young children already appear to have some grip upon the idea that they sometimes prefer one thing, while others around them often prefer something else entirely. Sometimes, like us, such children adopt a “live­and­let­live” attitude toward such differences. Again, like us, they are not always so tolerant, and are quick to brand such divergent tastes as wrong­headed, perverse, and sanctionable – as preferences that one is mistaken to have. Where the developmental story is in all of this, however, is far from obvious.

Step 2: Brute and other demi-sec facts

When it comes to knowledge of the material world, two familiar pos­sibilities compete for our attention. One of these, which, in most quar­ters, has fallen into professional disrepute, is that we are somehow gifted with the ability to directly process information naturally given off by the environing world. It is this theory of “immaculate percep­tion” (Chandler, 1988) that has been directly challenged by what is still called (some sixty years later) the “new look” in psychology – a once radical account bent upon demonstrating, for example, that while chil­dren of the wealthy commonly see ordinary pocket­change as no bigger than a mote in their eye, offspring of the poor routinely imagine the same coins to be altogether bigger than real life (Bruner and Goodman,

Michael J. Chandler and Travis Proulx 207

1947). Given such demonstrations, several generations of psychological experts have worked to make it plain that what people claim to know very often fails to reduce to any of the raw ballistical firings that natu­rally ricochet off of the material world – ideas that, more often than not, amount to some cross­product of what is on offer and what the world is imagined to be.

Although, like most contemporary adults, we personally subscribe to some version of the second and more constructivist of these accounts, this is hardly the point, at least insofar as our task, for the moment, is to get at whatever ordinarily varies in the naïve epistemologies of develop­ing persons. Rather, what does matter is whatever it is that young per­sons, of this age or that, ordinarily imagine knowing entails. What the available research suggests is that young persons (including a certain number of college­age respondents) seem more broadly committed to some copy­theory of knowledge than are most adults (Moses and Chandler, 1992).

Before getting too caught up in the details of how increasing maturity might impact on such early epistemic views, it will prove useful to first begin by attempting to distinguish between simpler and harder cases, and by pointing out (as Hammer and Elby, 2002, make clear) that it is not always more “sophisticated” to imagine that every case is necessar­ily one of the hard ones. In the first of these instances it is possible to identify a class of easy cases involving matters of direct sense percep­tion (e.g., “The cat is on the mat”) – knowledge claims that, were they to be called into question, would cause all of us to first doubt our own or others’ visual acuity or, worst still, the sanity of whoever it was that imagined things differently.

Not every case, however, is so simple. Of all the things we claim to know, for example, only some have come to us through simple direct perception. Rather more common (and this is especially true for young children, whose experience is necessarily limited) are all our putative claims to knowledge that we have learned indirectly from others who are judged to have had better reasons or better opportunities to know (what Hammer and his associates, e.g., Hammer and Elby, 2002, character­ize as merely “propagated stuff,” p. 12). We are, of course, occasionally misled by such rumors and second­hand offerings. Nevertheless, we cannot, and do not, ordinarily doubt much of what we are led to believe – that the earth, for example, is round, or that the heart pumps blood. We ordinarily take such “propagated stuff” at face value, presumably because, while we may not know about these things directly, we trust that, given access to the required evidence, we too (given appropri­ate educational opportunities) would arrive at similar conclusions. Very

Stalking young persons’ changing beliefs208

young children have a large store of such hand­me­down knowledge, and are often quick and confident to assert all sorts of things, simply because their mother or their teacher “told them.” Given such pros­pects, educators are put on special notice that, however much they may otherwise feel despised, owning an academic post is too often seen as the same thing as having a special corner on the truth. While working out how such knowledge claims can sometimes go wrong can take a lifetime, there is no reason to suppose that the beliefs that we come to in such indirect ways automatically demand that we radically restructure our earliest understanding about what the business of understanding brute facts is taken to be. Still other sorts of knowledge (and here we rely again on Hammer and associates, e.g., Hammer and Elby, 2002) are imagined to arise from acts of simple “invention” (“How do you know your cat’s name is Mouser?” … “I made it up”) or (more compli­cated still) have as their imagined source still other ideas that form the basis for some inference (“I figured it out”).

Although, according to Hammer and his colleagues (2002), the pros­pect that we are all capable of “inventing” new knowledge ends up being listed out as only one more in a long train of possible knowledge resources, there are, nevertheless, good reasons to suppose that the act of envisioning such prospects amounts to something of a new quan­tum leap in what being knowledgeable could possibly mean. Unlike so­called “propagated stuff,” and knowledge owed to acts of direct sensory perception, such imaginings would seem to presuppose a model of the mind that is altogether more complex than that presupposed by passive copy­theories of meaning making (Chandler, 2001) – a model on which formal programs of education ordinarily rely. Our own research efforts (e.g., Carpendale and Chandler, 1996; Chandler and Lalonde, 1996) and that of others (e.g., Carpendale, 1995; Lalonde, 1996) would sug­gest that such derivative insights require something like an “interpre­tive theory of mind” that makes room for the possibility that persons are not simply the passive recipients of information impressed upon them from the material world, but, instead, recognize that we all play some active agentive role in generating the contents of our own and others’ mental lives – an accomplishment that our own research sug­gests is most commonly achieved in middle childhood. Knowledge of this sort, as we will go on to argue, fits better into the category of what we will term “social facts.”

Be all of this as it may, there would appear to be precious little that is currently known about children’s earliest commitments to some increas­ingly constructivistic or interpretive folk epistemology that would obli­gate us to suppose that they do (or should) take their beliefs about the

Michael J. Chandler and Travis Proulx 209

material world to be somehow illegitimate or tentative rather than cer­tain. Like adults, such children need have no special commitment to the prospect that, because some knowledge is clearly constructed, all perceptual knowledge is therefore automatically subject to doubt. “It is” (as Hammer and Elby, 2002, argue) “hardly sophisticated to con­sider it ‘tentative’ that the Earth is round, that the heart pumps blood, or that living organisms evolve (p. 186).”

We may, especially in the case of various sorts of “propagated” knowl­edge, sometimes end up being duped, but “there are [of course] worse things that can happen to a man [sic] in this world than being duped” (Chandler, 1987). On this final point we mean to take special issue with Kuhn and her colleagues (2002), who offer up suspect evidence meant to demonstrate that, more than half of the time, even among suppos­edly “sophisticated” college students, such young adults are tentative and uncertain about their grip upon even the material world.

Step 3: Matters of social fact and value

So far in this chapter we have worked to set two broad epistemological categories – one containing beliefs that are externally justified and gen­erally agreed upon, the other contested and treated as fancy. We have also argued that passably grown­up insights about the nature of beliefs about both brute facts and matters of aesthetic preference are ordi­narily available to young persons from a very early age. The same, we mean to show, cannot also be said about our early grip upon social facts and shared values. Our common problem with facts belonging to this remaining quarter, we want to suggest, arises because social facts and values are unlike matters of direct sense perception (or similarly “prop­agated stuff”), at least insofar as they boast no external referent, and differ from matters of taste in that they are commonly shared. Work­ing out how to relate to beliefs that present this mixed profile evidently takes some special doing – a task that, because it apparently eats up a great deal of ontogenetic time, supports a real developmental story.

The general thrust of our argument is that once confronted with anything that might later qualify as a bona fide instance of some real social fact of shared value, the first strategic move undertaken by young persons (and far too many titular adults) is to work to empty out this new awkwardly looming class of truth claims into one or the other of their already better understood categories of sense data or personal preferences. Consequently, much that is later destined to qualify as institutional facts and values are initially discounted as either another example of God­given brute facticity or, alternatively, written off as

Stalking young persons’ changing beliefs210

mere subjective musings. The commonplace character of such “out­sourcing” operations has, no doubt, contributed to the once elaborate literature on so­called “childhood realism” (Wellman, 1990). As such, no one was ever surprised to learn that “thou shalt brush thy teeth up and down” is regularly taken by young school­age children to be some­thing like the eleventh commandment, chipped in stone for all eternity, or to learn that, well into their middle­school years (Nucci and Nucci, 1982), young persons are ready to believe that “Thou shall not kill” and “Thou should not wear one’s pajamas to school” are objective laws of a similar caliber. Similarly, deciding that some knowledge claims have no better warrant than blind liking (mere matters of arbitrary opinion and personal taste) while others (e.g., growing children need between eight and ten hours of sleep) are rooted in better authority is a discov­ery that often requires most of adolescence to achieve. As these later examples are meant to show, it may well be the case that more “sophis­tication” is required to drive a proper wedge between social facts and mere personal opinions than is necessary to intuit the possibility that every agreed upon fact is not necessarily a brute fact. What already seems clear enough, however, is that much of childhood can be used up working it out that, when all is said and done, brute facts and arbitrary matters of taste are not the only serious contenders for serious epistemic thought.

Despite the growing prospect that working out what should and should not count as a brute fact, on the one hand, and what legitimately qualifies as an arbitrary matter of personal taste, on the other, might prove to have slightly different developmental arcs, and otherwise seems as different as night and day, it would still seem to be the case that both are what Gadamer (1960/1982) called “secret sharers,” all for the rea­son that “both are predicated on the same pernicious assumption that without absolute certainty everything is lost” (p. 28).

There are, then, what would appear to be good reasons to suppose that many of the variations in “epistemic stance” documented by inves­tigators concerned with the folk­epistemologies of adolescents and col­lege­age youth amount to variations on the theme of how best to offload the problems of social facts and values into the less troubled categories of brute facts and mere opinions. The concepts of “objectivism” and “dogmatism” that are regularly employed in the personal epistemol­ogy literature (Boyes and Chandler, 1992) appear, for example, to be notions aimed at capturing the familiar efforts of young people to turn institutional facts and values into less contested matters of brute fact. Similarly, the notions of radical skepticism, rampant subjectivism, and know­nothing nihilism are commonly used to capture what Marcia

Michael J. Chandler and Travis Proulx 211

(1966) has summed up as the “moratorium” period, and would seem to describe how adolescents and others work to discount social facts and values by re­assigning them to some unassailable category of pure taste.

When set in contrast to both matters of direct sense perception and personal tastes, our common task of coming to a mature understanding of social facts and shared values is, we suggest, altogether more com­plex and interesting. A rather long list of special circumstances would appear to conspire to make this true.

Some part of this greater task­complexity is owed, no doubt, to the heteronomous nature of young children’s early life experiences. In their roles as adult care­givers, teachers and parents ordinarily lay claim to knowing a great many things that, at least in the beginning, children are naturally obliged to take as true. Especially with regard to mat­ters of brute facticity, such hand­me­down knowledge (e.g., that stoves are hot, or that soap is not for eating) is often as good as it gets. The same cannot always be said, however, about matters of social facts and values. One of Piaget’s (1985) most important contributions, it would seem, was his insight that children’s progress toward moral maturity is heavily dependent on conditions of social equality common among peers, but hard to reproduce in the authoritarian context demanded of early parent–child interactions.

A second apparent obstacle to children’s early access to mature insights into the character of social facts arises because such under­standing is often heavily dependent on an appreciation of the situ­ated character of social­moral matters. Although there is continuing debate about precisely when in the course of their development young persons first acquire the rudimentary cognitive capacities required to process so­called “representational diversity” (Chandler and Carpen­dale, 1994), there is widespread agreement that children are not born into the world with such skills fully f ledged. Familiar evidence from the so­called “theories of mind” literature – evidence that rests per­haps too exclusively on one or another version of Wimmer and Pern­er’s (1983) now classic false belief tasks – is commonly interpreted as demonstrating that children as young as three or four already have an understanding that others may reach different conclusions concern­ing the “same” event. On the strength of such findings, some (e.g., Perner, 1991) have argued that preschoolers evidence an appreciation of epistemic matters that, while perhaps different in degree, are not different in kind from that subscribed to by older children and adults. Others, ourselves included (e.g., Chandler and Lalonde, 1996), have militated for a rather later start­up date, pointing out that all that

Stalking young persons’ changing beliefs212

is required to pass standard false belief tasks is an appreciation of the fact that those left in ignorance (e.g., poor benighted Maxi) hold to different beliefs than those (e.g., Maxi’s mother) who are better informed. None of this, it can be argued, has much to do with eventu­ally recognizing that even those with access to all of the same details still regularly fail to come to common beliefs. Our own data (e.g., Chandler and Carpendale, 1994; Chandler and Lalonde, 1996) sug­gest that children are typically six going on eight before first showing any beginning appreciation of the fact that there is an ineluctably sub­jective component to all acts of knowing. However long it may take for growing persons to acquire those cognitive competencies required for a sophisticated reading of situations involving representational diver­sity, it seems clear enough that the opening price paid for missing out on some or all of these capacities is necessarily higher when negotiat­ing social matters.

Similarly, it is not all that surprising that, as young people and the horizons of their social­cultural life both grow, so too do opportunities for coming to a more mature understanding of the more or less relativ­ized character of social facts and values. For much the same reasons, the social­moral lives of adolescents and young adults also optimally broaden, exposing them to the often tangled teleological hierarchies that dictate that the obeying of some maxims may come at the expense of others – “Can I lie to the Gestapo about the Jews hiding in my base­ment?” Eventually, we interact with others who never seem to share, who get ahead by cheating, who don’t like the colour of our skin, or value freedom over equality. How do great minds adjudicate these kinds of competing knowledge claims? How do ten­, or fifteen­, or twenty­year­olds?

Historical change too must play its part. According to Fromm (1941/1965), for example, the moral consensus of the deeply religious medieval world is widely understood to have begun to rupture dur­ing the Renaissance, and then splinter apart during the Enlighten­ment that followed. Where once our values were handed down from a raised dais, Voltaire (Milza, 2007) was calling for us to burn it all down and start again. The generations that followed have worked to pick up the pieces and carry on, much as do contemporary adults who, after the moral consensus of early childhood is finally ruptured in adolescence, seem to find a way to commit to a set of values, while acknowledging that these values may not be self­evidently true for all ages and individuals. As some of these ideas suggest, matters of value – both what we should want and how we should get it – have been the unacknowledged, primary focus of the epistemic literature

Michael J. Chandler and Travis Proulx 213

to date, for the simple reason, we suggest, that this is where epistemic development most commonly moves from objectivism, to subjective relativism, and ultimately to some construction of a serviceable mid­dle ground.

Finally, and in light of all that has just been said, it again hardly seems surprising that many of the changes observed in the personal epistemologies of adolescents and young adults have been shown to be especially marked among those privileged with expensive secondary and post­secondary liberal arts educations (Hofer and Pintrich, 1997; King and Kitchner, 1994; Perry, 1970; Reich et al., 1994).

Although the full details of this developmental story remain to be told, it is the case, we want to argue, that there really is an interesting, but untold, developmental story here. While we have no special way of imagining this fuller account, there are, we think, some caution­ary tales to be told about how to best keep the wheels of this shared task from coming spectacularly off. The balance of this short chapter involves a short listing out of such cautions.

Epistemic development everywhere or nowhere

We began this chapter with a long dirge lamenting the sad fact that the community of experts committed to the study of personal epis­temologies seems to agree on almost nothing. Where we mean to have come in the preceding pages is to some conceptual place from which the troubling reasons behind such radical disagreements are both more apparent and more easily avoided. Our own position, as we have worked to make plain, is that the stuff about which knowledge claims are ordinarily made is not all stuff of the same kind or caliber. Rather, we claim to know certain things (brute facts, if you will) owed to our own (and others’) perceptual experiences, while other things amount to more personalized knowledge concerning our own and oth­ers’ aesthetic experiences. On top of both of these, we also worked to draw attention to the worlds of social facts and shared values that fit awkwardly into one or the other of the other available categories of material and aesthetic knowledge. More pointedly, we argued that epistemic development, when it occurs, standardly occurs in only the last of these categories – the one involving institutional facts and val­ues. If something like this is true, and if many of our colleagues focus their research attention on some but not others of these potential tar­gets of epistemic activity, then reasons for the wide­spread disagree­ments that have appeared in the ranks become more understandable and even predictable.

Stalking young persons’ changing beliefs214

For example, Hammer and his colleagues (2002), concerned as they are with how best to teach physics, have largely focused attention on what young persons think about matters of, and rumors about, brute facts. Not surprisingly, their data leads them to the all but necessary conclusion that there is little in the way of evidence for large, sweeping, across­the­board changes in epistemic reasoning. Kuhn and her col­leagues (2000), by contrast, begin with what promises to be a separate accounting of how young people think, not only about physical facts, but also about social facts and matters of taste. Given this promising beginning one might have (or better, we might have) anticipated that their conclusions would be something like our own – that is, that the bulk of real development is owed to changes in the arenas of social facts and values. Awkwardly, however, this is not what they report. Rather, the data they present seem to suggest an across­the­board pattern of yoked progress toward some increasingly “evaluative” stance in all of these areas of epistemic functioning. It is not obvious (given our own account of what brute facts and personal preference are standardly taken to mean) how something like this could possibly be. What would it actually look like, for example, to adopt an “evaluative” stance toward claims such as “Stoves are hot” or “Chocolate is better than vanilla”? The answer to this paradox, we are asked to suppose, is to be found in the kinds of items that Kuhn and her colleagues (2002) regularly use to assess young peoples’ thoughts about physical facts, social facts, and matters of taste – choices, we suggest, that are not sufficiently consid­ered. It is difficult to imagine, for example, how asking young people to adjudicate disagreements over explanations of brain function, atomic phenomena, or math problems actually involve judgments of “fact about the physical world” per se. Naturally enough, hypothetical disagree­ments over matters of sense perception or direct measurement would make poor questionnaire items, all for the reason that people almost never disagree about such matters, and suggesting that they do would lack credibility (e.g., Suzy believes the toy is round; Timmy believes the toy is square). The temptation, unfortunately, appears to have been to choose test items – even for physical facts – that avoid physical facts altogether, all in an effort to present competing knowledge claims about which people may plausibly disagree. Of course, this seriously begs the question. If Kuhn and Weinstock (2002) were to operationalize “physi­cal facts” in ways that emphasized competing knowledge claims that actually fit the category, we imagine that greater light would have fallen between epistemic development in matters of fact and value, with objec­tive stances about physical matters remaining stubbornly in place well into young adulthood (Hallett et al., 2002).

Michael J. Chandler and Travis Proulx 215

Educational implications and recommendations for future research

Although we have no special expertise that would entitle us to seri­ous professional opinions about how, in light of the findings presented here, educators might best proceed in allying themselves with young people’s epistemic struggles, there are, perhaps, reasons to recommend that their efforts might be best spent in areas earmarked as involv­ing “social” facts and values. If available research findings are to be believed (Flavell et al., 1992; Moses and Chandler, 1992), young chil­dren acquire an appreciation of the inherently subjective character of strict matters of personal taste almost as soon as they learn to walk and talk. There may, of course, be ongoing difficulties in learning to extend such early insights into every possible quarter (e.g., few teenagers are prepared to believe that their parents are within their right to enjoy the antique music that they often seem to prefer), but, in principle, there would seem to be no formal impediment to their doing so.

Much the same also seems to be true of matters of “brute” fact, but for the opposite reason. Here the common environment seems to overtake the modest differences that set one person’s interpretive machinery apart from the next, and variability is generally lost. Say­ing that black is white is not, of course, unheard of, but it is rare, and seeks its explanation in other than the perceptual givens. This is not, of course, true when it comes to theorizing about such putative facts, and it is here that Kuhn and Weinstock (2002) succeed in finding evidence of age­graded differences in interpretive abilities. Theories about facts are not, of course, (brute) facts, in and of themselves, and so are better understood instead as matters of active human construc­tion, more properly belonging to the same realm as do other social or institutional facts. In short, when it comes to the “facts” themselves, something like “objectivism” or “naïve realism” will do nicely, and so there is little in the way of a real developmental story to be told here either.

So, it would appear, everything – or at least everything potentially educational – comes down to so­called “social” or “institutional” or value­impregnated facts. Here, there is good reason to suspect, educa­tion has the possibility of getting a real toehold. Exactly how old, or how well educated, were you, for example, when it first occurred to you that the gods of other peoples were not so different from your own gods? Or, at what point in your development did it first seem credible that a union between a man and a woman was not the only sort of union that might deserve our compassion and respect?

Stalking young persons’ changing beliefs216

None of the above is meant to suggest, however, that simply being alerted to the fact that, in different places and at different historical moments, different people have believed in different things, is anything like an end in itself. Quite to the contrary, such insights have had (as often as not) the disruptive effect of “poisoning” some earlier well of comfortable certainty, and ushering in a period of “epistemic loneli­ness” (Chandler, 1975, 1987) that cast young persons into a kind of know­nothing relativism, or period of “moratorium” (Marcia, 1966) equivalent to being unable to either keep to the surface or otherwise touch bottom. Where educators have been (e.g., Perry, 1970), and, no doubt, will continue to go on being, especially helpful (e.g., King and Kitchner, 2002) is in allying themselves with the struggles of young persons to achieve some better understanding of the inherently subjec­tive, cultural­bound, and historical character of what we hold out to be self­evident truths. What is probably the best help that the research community can offer up to assist educators in this process is to aid in narrowing and sharpening their impossibly large task by providing information about where young people’s doubts are felt most acutely. For the moment, the answer to this question appears to be in the domain of social and institutional and value­impregnated claims to truth.

Conclusion

What we have attempted to argue over the course of this chapter is that: (a) most work dealing with epistemic development has made few if any distinctions among the kinds of knowledge over which competing knowledge claims generally compete; and (b) that by not making these distinctions, important differences concerning the direction and scope of epistemic development are concealed. What we have attempted here amounts to an effort to introduce such distinctions after­the­fact, all in the hope of un­snarling what has become a rat’s nest of competing claims about competing knowledge claims. Our own short conclusion at the end of this effort is that, if epistemic development is to be found, it will be found in the contested areas of social facts and shared values, and not in the realms of young people’s developing understanding of brute facts or matters of personal taste.

R EF ER ENCES

Baxter Magolda, M. B. (1992). Knowing and reasoning in college: Gender-related patterns in students’ intellectual development. San Francisco: Jossey­Bass.

Michael J. Chandler and Travis Proulx 217

Boyes, M. and Chandler, M. J. (1992). Cognitive development, epistemic doubt, and identity formation in adolescence. Journal of Youth and Adoles-cence, 21(3), 277–304.

Bruner, J. and Goodman, C. C. (1947). Value and need as organizing factors in perception. Journal of Abnormal and Social Psychology, 42, 33–44.

Carpendale, J. I. M. (1995). On the distinction between false belief understanding and the acquisition of an interpretive theory of mind. Unpublished doctoral dissertation, University of British Columbia, Vancouver, BC.

Carpendale, J. I. M. and Chandler, M. J. (1996). On the distinction between false belief understanding and subscribing to an interpretive theory of mind. Child Development, 67(4), 1686–706.

Chandler, M. J. (1975). Relativism and the problem of epistemological loneli­ness. Human Development, 18(3), 171–80.

(1987). The Othello effect: Essay on the emergence and eclipse of skeptical doubt. Human Development, 30, 137–59.

(1988). Doubt and developing theories of mind. In J. W. Astington et al. (Eds.), Developing theories of mind (387–413). New York: Cambridge Uni­versity Press.

(2001). Perspective taking in the aftermath of theory­theory and the collapse of the social role­taking enterprise. In A. Tryphon and J. Voneche (Eds.), Working with Piaget: Essays in honour of Barbel Inhelder (39–63). London: Psychology Press.

Chandler, M. J. and Carpendale, J. I. M. (1994). Concerning the rumored fall­ing to earth of Time’s Arrow. Psychological Inquiry, 5, 245–8.

(1998). Inching toward a mature theory of mind. In M. Ferrari and R. Sternberg (Eds.), Self-awareness: Its nature and development (148–90). New York: Guilford Publications.

Chandler, M. J., Hallett, D., and Sokol, B. W. (2002). Competing claims about competing knowledge claims. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (145–68). Mahwah, NJ: Lawrence Erlbaum Associates.

Chandler, M. J. and Lalonde, C. E. (1996). Shifting to an interpretive theory of mind: 5­ to 7­year­olds’ changing conceptions of mental life. In A. Sameroff and M. Haith (Eds.), The five to seven year shift: The age of reason and responsibility (111–39). Chicago, IL: University of Chicago Press.

Chandler, M. J. and Sokol, B. W. (1999). Representation once removed: Chil­dren’s developing conceptions of representational life. In I. E. Sigel (Ed.), Development of mental representation: Theories and applications (201–30). Mahwah, NJ: Lawrence Erlbaum Associates.

Chandler, M. J., Sokol, B. W., and Wainryb, C. (2000). Beliefs about truth and beliefs about rightness. Child Development, 71(1), 91–7.

Elgin, C. (1989). The relativity of the fact and the objectivity of value. In M. Krausz (Ed.), Relativism: Interpretation and confrontation (86–98). Notre Dame, IN: University of Notre Dame Press.

Flavell, J. H., Flavell, E. R., Green, F. L., and Moses, L. J. (1992). Young chil­dren’s understanding of different types of beliefs. Child Development, 61, 915–28.

Stalking young persons’ changing beliefs218

Fromm, E. (1941/1965). Escape from freedom. New York: Avon Books.Gadamer, H. G. (1960/1982). Truth and method. New York: Seabury Press.Hallett, D., Chandler, M. J., and Krettenauer, T. (2002). Disentangling the

course of epistemic development: Parsing knowledge by epistemic con­tent. New Ideas in Psychology, 20, 285–307.

Hammer, D. and Elby, A. (2002). On the form of a personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (169–90). Mahwah, NJ: Lawrence Erlbaum Associates.

Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learn­ing. Review of Educational Research, 67(1), 88–140.

King, P. M. and Kitchener, K. S. (1994). Developing reflective judgment: Under-standing and promoting intellectual growth and critical thinking in adolescents and adults. San Francisco: Jossey­Bass.

(2002). The reflective judgment model: Twenty years of research on epis­temic cognition. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing (37–61). Mahwah, NJ: Lawrence Erlbaum Associates.

Kitchener, K. S. and King, P. M. (1981). Reflective judgment: Concepts of jus­tification and their relationship to age and education. Journal of Applied Developmental Psychology, 2, 89–116.

Kuhn, D. and Weinstock, M. (2002). The development of epistemological understanding. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing (121–44). Mahwah, NJ: Lawrence Erlbaum Associates.

Lalonde, C. E. (1996). Children’s understanding of the interpretive nature of the mind. Unpublished doctoral dissertation, University of British Columbia, Vancouver, BC.

Marcia, J. E. (1966). Development and validation of ego identity status. Journal of Personality and Social Psychology, 3, 551–8.

Milza, P. (2007). Voltaire. Paris: Perrin.Moses, L. J. and Chandler, M. J. (1992). Traveler’s guide to children’s theories

of mind. Psychological Inquiry, 3, 286–301.Nucci, L. P. and Nucci, M. S. (1982). Children’s social interactions in the con­

text of moral and conventional transgressions. Child Development, 53(2), 403–12.

Perner, J. (1991). Understanding the representational mind. Cambridge, MA: MIT Press.

Perry, W. G. (1970). Forms of intellectual and ethical development in the college years: A scheme. New York: Holt, Rinehart and Winston.

Piaget, J. (1985). The equilibration of cognitive structures. Chicago: University of Chicago Press.

Pillow, B. H. (1991). Children’s understanding of biased social cognition. Developmental Psychology, 27(4), 539–51.

Michael J. Chandler and Travis Proulx 219

(1999). Epistemological development in adolescence and adulthood: A mul­tidimensional framework. Genetic, Social, and General Psychology Mono-graphs, 125(4), 413–32.

Reich, K. H., Oser, F. K., and Valentin, P. (1994). Knowing why I now know better: Children’s and youth’s explanations of their worldview changes. Journal of Research on Adolescence, 4(l), 151–73.

Repacholi, B. M. and Gopnik, A. (1997). Early reasoning about desires: Evi­dence from 14­ and 18­month­olds. Developmental Psychology, 33(1), 12–21.

Rorty, R. (1991). Inquiry as recontextualization: An anti­dualist account of interpretation. In D. R. Hiley et al. (Eds.), The interpretive turn: Philosophy, science, culture (59–80). Ithaca, NY: Cornell University Press.

Schommer, M. (1990). Effects of beliefs about the nature of knowledge on comprehension. Journal of Educational Psychology, 85(3), 498–504.

Searle, J. R. (1983). Intentionality: An essay in the philosophy of mind. Cam­bridge: Cambridge University Press.

Wellman, H. M. (1990). The child’s theory of mind. Cambridge, MA: MIT Press.

Wimmer, H. and Perner, J. (1983). Beliefs about beliefs: Representation and constraining function of wrong beliefs in young children’s understanding of deception. Cognition, 13, 103–28.

220

8 Epistemological development in very young knowers

Leah K. Wildenger Syracuse University

Barbara K. Hofer Middlebury College

Jean E. Burr Hamilton College

Personal epistemology is a rapidly advancing area of research in the field of psychology concerned with the nature of human knowledge. Although epistemology has historically been pertinent to philosophers, it has more recently been adopted by psychologists concerned with how people develop, interpret, evaluate, and justify knowledge and beliefs about knowing. An understanding of epistemological development has significant implications for the ways individuals consider and approach the process of knowing and learning in a wide range of contexts, across the life span. Although much of the existing research has been focused on college students, recent work has expanded to include younger knowers and to bridge the theoretical gaps between epistemological understand­ing and other cognitive developmental processes, such as theory of mind. This chapter addresses the educational context of young knowers and provides research on the origins of epistemological development, with a description of two studies from our lab, one exploring epistemology and theory of mind, and the other domain specificity of epistemological development. We conclude with methodological recommendations for other researchers, suggestions for future research, and implications for educational practice.

The context of early childhood education

Although our work together has primarily been focused on basic research in expanding the developmental framework of personal epis­temology to include young children, we also find the context in which young knowers develop to be of importance. Variability is a constant theme in early childhood education for three­ and four­year­olds. Some children attend structured preschools, others attend daycare centers,

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 221

some attend family daycare in another person’s home, and still others remain in their own homes with a family member or hired caregiver.

Across these settings, children’s experiences are diverse (Zigler and Finn­Stevenson, 2007). The National Association for the Education of Young Children (NAEYC) promotes specific, research­based guide­lines for developmentally appropriate practice (DAP) in early childhood programs (Bredekamp and Copple, 1997) and offers a national accredi­tation system for programs that meet these standards. DAP involves: (1) creating a community of learners; (2) teaching to enhance develop­ment and learning; (3) constructing appropriate curriculum that inte­grates across a variety of disciples that are socially relevant, intellectually engaging, and personally meaningful to the children; (4) assessing children’s learning and development using authentic measures; and (5) establishing reciprocal relationships with families. Unfortunately, large­scale evaluations of early childhood settings reveal that most chil­dren (53 percent) who are enrolled in preschool and childcare programs in the US receive poor or inadequate care, relative to these types of standards (National Institute of Child Health and Human Development (NICHD) Early Care Research Network, 2005). As a result, there is great variability in children’s readiness for school, particularly across racial and ethnic groups (Brooks­Gunn et al., 2007).

Historically, kindergarten was viewed as a bridge that allowed chil­dren the opportunity to adjust to being in a school environment before they faced the academic rigor of the subsequent elementary grades. As a result, most kindergarten curricula focused on supporting social, cog­nitive, physical, and emotional development, a logical extension of DAP (Goldstein, 2007; Graue, 2001). However, the recent adoption of the no child left behind legislation in the US (and its associated third­grade standards) has caused many states to implement academic benchmarks for the kindergarten year (Goldstein, 2007). Consequently, kindergar­ten is facing an increase in its academic focus. This can be seen in the push to move from half­ to full­day programs (DeCicca, 2007) and increasing attention to teaching specific standards (Hatch, 2002). Many kindergarten teachers continue to struggle to balance kindergarten’s important historical functions with these new academic requirements (Goldstein, 2007).

This shift from preschool into an academically rigorous kindergarten likely affects children’s learning and understanding of school. Rather than being allowed to create their own knowledge through exploration of personally relevant topics, as they did in preschool, many kindergarten children are presented with specific pieces of knowledge that they are

Epistemological development in very young knowers222

“required” to know. It logically follows that this could lead children to question what they and others know and influence their understanding of the process by which we gather knowledge. Accordingly, understand­ing the context of this academic transition is important for researchers examining children’s changing conceptions of knowledge and knowing. Similarly, early educators may benefit from learning more about the trajectory of epistemological development in young children.

Personal epistemology as a cognitive developmental process

Research on personal epistemology has typically neglected young chil­dren. The foundations of personal epistemology as a pursuit of the discipline of psychology can be traced to Perry (1970), whose longi­tudinal work with Harvard undergraduates resulted in a scheme of intellectual development during the college years. There are several stage­based, developmental models of personal epistemology (Baxter Magolda, 1992; Belenky et al., 1986; King and Kitchener, 1994; Kuhn, 1991; Perry, 1970). The trajectories of these models suggest a general transition from a dualistic perspective of knowledge to a more relativ­istic stance and ultimately to a contextual, constructivist perspective on knowing (Hofer and Pintrich, 1997). In most schemes, individuals begin in a period termed either dualism or absolutism, with a concrete view of knowledge as strictly correct or incorrect, and hence the belief that truth is completely discernable. Gradual movement into a period typically labeled multiplism then occurs, when uncertainty about knowledge begins to prevail. Individuals acknowledge the existence and equal merit of multiple perspectives. A final transition into a stage identified as evaluativism by Kuhn (1991) occurs when individuals see knowledge as constructed, and are able to weigh conflicting viewpoints and evaluate them in light of relevant evidence rather than uniformly acknowledging their value. Justification of knowledge also becomes crucial in this stage.

At the core of each of these stages is the centrality of subjective versus objective states of knowledge; Kuhn and Weinstock (2002) suggest that underlying the transitions to increasing epistemological understanding is an individual’s gradual integration of the subjective and objective dimensions of knowledge. Kuhn and Weinstock weave their analysis of the interaction of these two dimensions into a generally accepted epistemological framework: the absolutist sees knowledge from an objective perspective, the multiplist takes a subjective view, and finally, the evaluativist achieves a mature balance of the two, coordinating a

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 223

personal and subjective frame of knowing with an awareness of how knowledge can be verified.

These stages identified by researchers reflect research with individu­als in adolescence and adulthood, with little attention to epistemologi­cal states in childhood. Many researchers in this field have expressed concern about the lack of knowledge surrounding the developmental origins of personal epistemology (Burr and Hofer, 2002; Chandler and Carpendale, 1998; Chandler et al., 2002; Kuhn and Weinstock, 2002) and have underscored the need for additional research with younger participants. An unpublished study by Mansfield and Clinchy (1985) appears to be the first attempt to explore the epistemological develop­ment of young children, and only recently has this work been further pursued by others.

Theory of mind

Although young children have not historically been a focus of research on personal epistemology, there has been an abundance of research on development of the cognitive construct called “theory of mind” with younger populations. Accomplishment of a theory of mind typically occurs in children sometime between the ages of three and five, dur­ing the preschool years. This developmental achievement, a well­rep­licated and widely accepted finding in empirical literature (Wellman et al., 2001), is characterized by the child’s ability to understand three domains of mental life: that the mind exists, that it can have different states and processes, and that causality exists between mental proc­esses and actions (Lee and Homer, 1999). Presence or absence of this construct is typically assessed through a standard “false­belief task.” A child’s understanding of the concept of false belief indicates that he/she recognizes that mental states are distinct, and perhaps different, from reality. The fact that three­year­olds typically fail false­belief tasks, whereas five­year­olds are typically successful, provides com­pelling evidence for a developmental hypothesis that theory of mind changes dramatically during the preschool years to include a represen­tational understanding of the mind (Wellman et al., 2001).

In a typical false­belief task, a child is invited to describe the contents of a familiar candy box, without physically looking inside. The experi­menter shows the child that, contrary to her expectations, the box actu­ally contains pencils. The child is then asked to describe what a friend would think is in the box, upon entering the room. Prior to achievement of a theory of mind, the child would assert that a friend would expect pencils rather than candy, ignoring the fact that she has had access

Epistemological development in very young knowers224

to additional information not available to the friend. She would also typically deny that she herself had expected to find candy in the box only a few moments ago (Flavell, 1999). In contrast, a child with a theory of mind would anticipate a false belief and assert that the friend would also expect candy, as well as acknowledge her own false belief. In their meta­analysis of theory of mind literature, Wellman et al., (2001) found consistent developmental patterns in preschoolers across varied countries and task manipulations for false­belief understanding in 178 empirical studies representing upwards of 4,000 children.

Connections between personal epistemology and theory of mind

As the field of personal epistemology has expanded in scope, research­ers have attempted to locate the construct in a broader developmental context (Hofer, 2001), as well as to expand the trajectory of develop­ment. Theory of mind tasks relate to epistemological understanding in that they require the separation and understanding of the subjective and objective components of knowledge. Children must fully grasp that people can form different mental representations of the same situations or objects, an awareness that enables movement away from egocen­trism. This recognition that individual subjective knowledge states can differ is a central component of epistemological development.

Theoretical origins of the distinction between subjective versus objec­tive knowledge states, integral to discussions of epistemological develop­ment, are evident in the research of Piaget (1929/1951), who defined the young, preschool­aged, preoperational child as exclusively “realistic … ignoring the existence of self and thence regarding one’s own perspec­tive as immediately objective and absolute” (Piaget 1929/1951, p. 34). In his writings on childhood egocentrism, Piaget contrasts this state of realism with a more advanced state that he deems “objectivity,” in which one becomes aware of “the countless intrusions of the self in eve­ryday thought and the countless illusions which result” (p. 34). Piaget suggested that egocentric children are cognitively unable to distinguish between these two states of knowledge. From an epistemological per­spective, this suggests that they are at a primitive level of epistemologi­cal reasoning. In contrast, slightly older children aged seven to eight, in Mansfield and Clinchy’s (1985) study, recognized that differences of opinion (i.e., whether a new TV show is good or bad) can originate in the person rather than the world, suggesting a degree of separation between subjectivity and objectivity. Clearly, some form of epistemo­logical development begins during childhood.

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 225

In many key respects, research on personal epistemology as investi­gated with older individuals and research on childhood theory of mind appears to involve examination of related cognitive activities (Hofer and Pintrich, 1997). Consequently, the two constructs have been theo­retically and empirically linked in recent years. Where childhood epis­temology is concerned, the most important aspect of theory of mind research is the child’s theory of belief (Kitchener, 2002). A critical aspect of theory of mind is a rudimentary understanding of the ways in which individuals’ beliefs and desires relate to and motivate their actions. Kuhn (2000) argues that the core conceptual link between the two constructs is that successful performance in a false­belief task, evi­dence of a developed theory of mind, requires that a child understand that a person’s claims are actually their beliefs about the state of things, a result of the process of knowing. Therefore, the development of theory of mind is closely linked to epistemological awareness.

Despite the critical links between personal epistemology and theory of mind, it is important to recognize that each is distinct. Theory of mind ability to understand the relations among others’ beliefs, actions, and desires appears dependent on a basic understanding of the nature of knowledge, whereas epistemological understanding is considerably broader, encompassing a full range of conceptions regarding knowl­edge and knowing that develop over a broader period of time and into adulthood. Including theory of mind in the developmental trajectory of personal epistemology is helpful in further illuminating the ways that young children think about knowledge and knowing.

Epistemological development in children prior to the onset of absolutism

Several researchers have made theoretical proposals for preliminary stages in childhood epistemological development preceding the abso­lutist or dualist stance, which has generally been empirically identified as the earliest stage, albeit largely in studies conducted with college students. Perry (1970) compared the black­and­white distinctions made by dualists to the way in which small children divide their world in terms of “family and the vague inchoate outside” (p. 66). He suggested, therefore, that small children operate at a pre­dualist epistemological state. In the original scheme of King and Kitchener (1994), the first stage of reflective judgment is characterized by an amorphous distinc­tion between knowledge and beliefs, and seems more primitive than dualism. These researchers rarely found evidence of this type of think­ing in early high school students, and hypothesized that it was more

Epistemological development in very young knowers226

characteristic of young children, although they did not test younger participants. Kuhn and Weinstock (2002) re­evaluated assumptions of their original scheme in more recent work, proposing that a realist stage exists prior to absolutism. A realist accepts blindly, needs no justifi­cation for knowledge, and evaluates statements for correctness; again, this was hypothesized but not tested empirically. Chandler et al. (2002) similarly hypothesize a primitive stage called “naïve realism” in very young children. Thus there is a convergence of propositions that an epistemological state preceding dualism most likely exists in children younger than those tested in typical studies of personal epistemology, but in the absence of empirical evidence, there has been little consensus on its characteristics.

Our first study of very young knowers: theory of mind and epistemology

In our initial study investigating theory of mind and personal episte­mology, the third author conducted interviews with three­ to five­year­olds, and results led us to posit a critical link that facilitates the mutual conceptualization of these two constructs (see Burr and Hofer, 2002, for full results of this study). We identified an association between a dualistic stance, in which a clear “right” and “wrong” may be per­ceived, and the ability to understand that others’ beliefs may differ from the child’s own due to dissimilar experiences, via theory of mind. Thus, we argue that theory of mind is a cognitive achievement accompanied by a transition from a primitive precursor of dualism to dualism itself. Revisiting Mansfield and Clinchy’s (1985) descriptions of epistemo­logical understanding in three­ and four­year­old children, who were classified as absolutists, we draw differing conclusions. Based on the children’s tendency to justify knowledge with personal experience as well as their marked anxiety in response to disagreement, we claimed that the children’s lack of distinction between knowledge and belief states instead warrants the label “pre­dualist” (Burr and Hofer, 2002).

Expanding the developmental trajectory: egocentric subjectivity

Based on our finding of significant correlations between maturity of epistemological understanding and theory of mind capability, we (Burr and Hofer, 2002) proposed a re­vamped early developmental trajec­tory. (See Figure 8.1 for an overview of epistemological development in childhood and adolescence.) A state of pre­dualism or naïve realism

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 227

exists prior to theory of mind achievement. This is then followed by a transitional phase in which a primitive form of epistemological under­standing exists despite the continued absence of theory of mind. The subsequent achievement of theory of mind coincides with the arrival of a dualistic or absolutist epistemological stance. The nature of the transitional state suggests that there is an epistemological level that pre­cedes theory of mind development.

Within our early research on young children, we also focused on the coordination of subjective and objective components of knowing as a critical task underlying epistemological development, and propose

Early childhood0–3

Childhood3–7

Pre-adolescence7–12

Adolescence12–18

Young adult18–25

Perry (1970)

Dualism Multiplicity

Contextual relativism Commitmentwithin relativism

King and Kitchener (1994)

(Pre-reflective Pre-reflective Quasi-reflective ---stage 1) stage 2 stage 3 and 4

Mansfield and Clinchy (1995)

Absolutists Objective multiplists Multiplists

Kuhn, Cheney, and Weinstock (2000)

(Realism) Absolutists Multiplists Evaluativists (not achieved by many; timeline varies)

Chandler et al. (2002)

(Naïve realism) Defended Dogmatism- Post-skepticalrealism skepticism rationalization

Burr and Hofer (2002)

Egocentric Predualism Dualism subjectivity No ToM ToM

No theory of mind (ToM)

Figure 8.1: Proposed timeline for epistemological development across childhood, with generalized ages at which stages are likely to occur. Stages that have been hypothesized but do not currently have empiri­cal support are indicated in parentheses. The dashed line indicates that additional stages occur later in life

Epistemological development in very young knowers228

that a stage of egocentric subjectivity should be added to the trajectory (see Figure 8.1). This new stage corresponds to the pre­dualist epis­temological stage, but more aptly depicts the subjectivity that occurs during the early preschool years. Children in this stage cannot distin­guish between personal knowledge and objective reality, assuming that they are one and the same. Because young children are egocentric, and believe that everyone shares their exact view of the world, the subjectiv­ity of pre­dualism is markedly different from the subjectivity character­izing the stage of multiplism identified in adolescents, for example Burr and Hofer, 2002.

Our second study of very young knowers: domain dependency and epistemology

Although it is clear that theory of mind onset marks a turning point in epistemological understanding, there is a dearth of research examin­ing the nuances of the development of epistemology during early and middle childhood. One area where there is a particular need for further research is domain dependency, one general explanation for discrepan­cies in research on early epistemological thinking (Kuhn et al., 2000; Kuhn and Weinstock, 2002). Many researchers posit that epistemologi­cal development may progress differently according to particular type, or domain, of judgment. Conceptualizing the development of domain­dependent epistemological thought as a process spanning the lifetime may help to clarify contradictory findings that identify “epistemological beginners” at multiple ages and developmental phases (Burr and Hofer, 2002), a concern raised most eloquently by Chandler et al. (2002).This might also permit better understanding of the educational implications of epistemological development in the preschool to grade twelve years.

“Domain” has been defined differently by various epistemology researchers (Hofer, 2006). For those working within the paradigm of epistemic beliefs, the word domain generally refers to academic disci­plines, such as math, history, or science (Muis et al., 2006).1 However, developmental psychologists have taken a different approach from those investigating epistemic beliefs, typically conceptualizing a domain as specific types of judgments. This line of research partitions the wider

1 We use the term “epistemic” to modify terms such as beliefs, as these are beliefs about knowledge (the epistemic), not beliefs about epistemology (Kitchener, 2002). Con­versely, we use “epistemological” to modify development, as this refers to the develop­ment of an individual’s personal epistemology (what counts as knowledge, where it resides, how it is justified, etc.), to related terms such as stages and positions, and to the processes that are a part of epistemological development.

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 229

concept of beliefs about knowledge states into various components such as beliefs about facts, values, or personal preferences (Kuhn et al., 2000; Mansfield and Clinchy, 2002; Mason et al., 2006; Wainryb et al., 2004). Although this definition has also led to a productive line of inquiry with older children and adolescents, this particular definition of domain is especially relevant to research with preschoolers, who are not yet likely to make explicit distinctions among academic subject areas.

Several compelling studies suggest that the timing of development differs depending on the specific type of belief under consideration. The division of the concept of “belief” into distinct domains serves to broaden the focus of research concerning the ways children think about beliefs, which has been criticized as restrictively narrow (Chan­dler and Carpendale, 1998), and relegated mainly to the realm of typi­cal false­belief tasks. Research addressing domain specificity of belief begins to rectify what Chandler and Carpendale (1998) label a glaring and problematic gap between issues of incontrovertible fact and highly subjective personal preference in this research. For an overview of the research on domain dependency in the epistemological development of children, see Table 8.1.

Kuhn and colleagues have asserted that the integration of subjective and objective knowledge follows a developmental trajectory that dif­fers by domain (Kuhn et al., 2000; Kuhn and Weinstock, 2002). They hypothesize that a transition from the absolutist to the multiplist stage would occur earlier within the more subjective domains of knowledge, such as personal preference and aesthetic judgments, as opposed to the relatively more objective domains of value and fact. Their vignette­based research with participants aged ten and over generally supports this trajectory (Kuhn et al., 2000; Kuhn and Weinstock, 2002).

In a longitudinal study of domain­dependent epistemological develop­ment in individuals ages ten to sixteen, Mansfield and Clinchy (2002) also presented their participants with vignettes. Characters disagreed about various issues, ranging from reconcilable matters of incontrovertible fact (i.e., “Do those clouds mean rain?”) to more amorphous, potentially irreconcilable issues of taste and value (i.e., “Who is the better artist?”). Mansfield and Clinchy used interviews (as opposed to the short­answer design utilized by Kuhn and colleagues) to discern reasons and explana­tions for judgments. Development from age ten to sixteen was associated with increasing “awareness of the complexity of both the outer world of objective ‘reality’ and the inner worlds of individual knowers, and a deeper understanding of the ways in which the two worlds intersect in the creation of knowledge” (p. 253). The observed correlations between age and increasingly flexible fact­opinion distinctions about issues in all

Tab

le 8

.1. P

rior

dom

ain-

depe

nden

cy r

esea

rch

on p

erso

nal e

pist

emol

ogy

with

chi

ldre

n

D

efin

itio

n of

dom

ain

Met

hodo

logy

Stu

dy

des

ign

and

age

of

part

icip

ants

Res

ult

s

Kuh

n et

al

., 20

00R

efer

s to

typ

es o

f ju

dgm

ents

in a

ran

ge

of d

omai

ns, f

rom

tho

se

cons

ider

ed h

ighl

y su

bjec

tive

to

high

ly

obje

ctiv

e. D

isti

ngui

shes

be

twee

n ju

dgm

ents

of

pers

onal

tas

te, a

esth

etic

s,

valu

es, t

ruth

abo

ut t

he

soci

al w

orld

, and

tru

th

abou

t th

e ph

ysic

al w

orld

.

Pre

sent

atio

n of

vig

nett

es

feat

urin

g ch

arac

ters

wit

h op

posi

ng c

laim

s in

the

for

m

“Rob

in s

ays

x, C

hris

say

s y.

” P

arti

cipa

nts

wer

e as

ked

(reg

ardi

ng t

he c

ompe

ting

cl

aim

s) “

whe

ther

onl

y on

e co

uld

be r

ight

” or

“w

heth

er

both

cou

ld h

ave

som

e ri

ghtn

ess.

” T

he in

terv

iew

was

co

nduc

ted

in a

sho

rt­a

nsw

er

form

at.

Cro

ss­s

ecti

onal

. Six

gro

ups

span

ning

mid

dle

child

hood

th

roug

h ad

ulth

ood:

fift

h­gr

ader

s, m

edia

n =

ten

yea

rs;

eigh

th­g

rade

rs, m

edia

n =

th

irte

en­y

ears

; tw

elft

h­gr

ader

s, m

edia

n =

sev

ente

en

year

s; u

nder

grad

uate

st

uden

ts, a

ge r

ange

eig

htee

n to

tw

enty

­one

. Gro

ups

five

and

six

wer

e ad

ults

fro

m

mid

tw

enti

es t

o la

te t

hirt

ies.

The

tra

nsit

ion

from

abs

olut

ism

to

mul

tipl

ism

pro

ceed

ed a

long

the

pr

opos

ed c

onti

nuum

of

dom

ains

; ea

rlie

st in

the

mos

t su

bjec

tive

and

la

test

in t

he m

ost

obje

ctiv

e ty

pes

of

judg

men

ts. H

owev

er, a

n ir

regu

lari

ty

in t

he d

ata

was

evi

dent

wit

hin

the

valu

e ca

tego

ry, w

here

abs

olut

ism

w

as m

ore

dom

inan

t an

d lin

geri

ng

than

was

ant

icip

ated

.

Man

sfiel

d an

d C

linch

y,

2002

Ref

ers

to t

ype

of ju

dgm

ent

conc

erni

ng is

sues

on

a ro

ugh

cont

inuu

m

from

mos

t ob

ject

ive

(an

imm

edia

tely

res

olva

ble

ques

tion

of

fact

) to

mos

t su

bjec

tive

(a

pote

ntia

lly

unre

solv

able

mat

ter

of

tast

e or

val

ue)

wit

h tw

o is

sues

con

tain

ing

a m

ore

bala

nced

mix

ture

of

subj

ecti

ve a

nd o

bjec

tive

co

mpo

nent

s in

the

mid

dle.

Vig

nett

es in

whi

ch

prot

agon

ists

dis

agre

e in

eac

h of

the

fou

r do

mai

ns w

ere

eith

er r

ead

or h

eard

by

the

part

icip

ants

, who

wer

e th

en

aske

d w

hy t

he p

rota

goni

sts

disa

gree

, whi

ch o

ne is

rig

ht,

and

whe

ther

the

dis

pute

can

be

res

olve

d, a

nd if

so,

how

. L

onge

r in

terv

iew

s st

rove

for

gr

eate

r de

pth

of a

naly

sis.

Lon

gitu

dina

l. P

arti

cipa

nts

test

ed a

t th

ree

poin

ts in

ti

me:

at

ages

ten

, thi

rtee

n,

and

sixt

een.

Obs

erve

d co

rrel

atio

ns b

etw

een

age

and

less

sha

rply

del

inea

ted

fact

­op

inio

n di

stin

ctio

ns a

bout

issu

es

in a

ll re

alm

s ex

cept

for

the

mos

t fa

ctua

l.

D

efin

itio

n of

dom

ain

Met

hodo

logy

Stu

dy

des

ign

and

age

of

part

icip

ants

Res

ult

s

Wai

nryb

et

al.,

200

1R

eflec

ts s

ubst

anti

ve

diff

eren

ces

amon

g re

alm

s of

soc

ial l

ife

and

soci

al t

houg

ht. F

inal

ca

tego

ry d

ivis

ions

: mor

al,

conv

enti

onal

, ps

ycho

logi

cal,

and

met

aphy

sica

l bel

iefs

.

Hyp

othe

tica

l cha

ract

ers

that

he

ld d

iver

gent

bel

iefs

tha

t w

ere

prev

ious

ly d

eter

min

ed

to b

e in

con

flict

wit

h th

e pa

rtic

ipan

ts’ o

wn

in

each

of

the

four

dom

ains

w

ere

desc

ribe

d. M

ulti

ple

asse

ssm

ents

of

part

icip

ants

’ at

titu

des

tow

ards

div

ersi

ty

of b

elie

f an

d th

eir

atti

tude

s to

war

d th

e di

sagr

eein

g ot

hers

w

ere

obta

ined

via

inte

rvie

w.

Cro

ss­s

ecti

onal

. Thr

ee

grou

ps: (

1) t

hird

­gra

ders

, m

ean

age

= e

ight

yea

rs,

nine

mon

ths;

(2)

sev

enth

­gr

ader

s, m

ean

age

=

thir

teen

yea

rs, t

wo

mon

ths;

(3

) co

llege

stu

dent

s, m

ean

age

= t

wen

ty­o

ne y

ears

, ten

m

onth

s.

Thi

rd­g

rade

rs m

ade

mor

e ne

gati

ve

judg

men

ts o

f di

verg

ent

belie

fs a

s co

mpa

red

to s

even

th­g

rade

rs a

nd

colle

ge s

tude

nts

in a

ll do

mai

ns

exce

pt m

oral

ity.

Acr

oss

ages

, pa

rtic

ipan

ts ju

dged

div

ersi

ty o

f m

oral

bel

iefs

to

be u

nacc

epta

ble.

P

arti

cipa

nts

in a

ll ag

e gr

oups

ge

nera

lly f

avor

ed u

nifo

rmit

y of

bel

ief

wit

hin

the

mor

al a

nd

conv

enti

onal

dom

ains

and

div

ersi

ty

of b

elie

f w

ithi

n th

e m

etap

hysi

cal

dom

ain.

Wai

nryb

et

al.,

200

4

Judg

men

ts o

f va

lue

and

fact

div

ided

into

qu

alit

ativ

ely

diff

eren

t ca

tego

ries

. Fin

al d

omai

n di

visi

ons:

mor

alit

y, t

aste

, fa

cts,

and

am

bigu

ous

fact

s.

Par

tici

pant

s w

ere

told

vi

gnet

tes

abou

t di

sagr

eein

g ch

arac

ters

onl

y on

e of

whi

ch,

acco

rdin

g to

a b

asel

ine

asse

ssm

ent,

sha

red

the

actu

al

view

of

the

part

icip

ant.

Ju

dgm

ents

of

rela

tivi

sm,

tole

ranc

e, a

nd t

he d

isag

reei

ng

pers

on w

ere

obta

ined

via

in

terv

iew

.

Cro

ss­s

ecti

onal

. Thr

ee

grou

ps: (

1) fi

ve­y

ear­

olds

; (2

) se

ven­

year

­old

s; (

3)

nine

­yea

r­ol

ds.

Fiv

e­ye

ar­o

lds

mad

e fe

wer

rel

ativ

e an

d to

lera

nt ju

dgm

ents

tha

n se

ven­

and

nin

e­ye

ar­o

lds

and

mor

e ne

gati

ve d

escr

ipti

ons

of d

isag

reei

ng

char

acte

rs. I

n th

e do

mai

ns o

f m

oral

and

fac

t di

sagr

eem

ents

, all

part

icip

ants

, reg

ardl

ess

of a

ge,

asse

rted

tha

t on

ly o

ne b

elie

f w

as

righ

t. H

owev

er, s

igni

fican

t ag

e di

ffer

ence

s w

ere

disc

erne

d in

ju

dgm

ents

of

the

mor

e su

bjec

tive

do

mai

ns o

f am

bigu

ous

fact

s an

d ta

ste.

Onl

y a

thir

d of

five

­yea

r­ol

ds

judg

ed m

ulti

ple

belie

fs t

o be

cor

rect

in

the

se d

omai

ns, i

n co

ntra

st t

o m

ost

of t

he s

even

­ an

d ni

ne­y

ear­

olds

. All

part

icip

ants

wer

e le

ast

tole

rant

of

dive

rgen

t m

oral

bel

iefs

.

Epistemological development in very young knowers232

realms except for the most factual provide additional support for Kuhn’s developmental continuum of domains.

Further support for the importance of examining domains separately is provided by Wainryb et al. (2001), who investigated thinking about diversity of beliefs with children and adolescents ranging in age from eight to twenty­two. The study also assessed participants’ attitudes toward the disagreeing person as well as opinions about the impor­tance of diversity of belief within specific domains. Participants were asked about a hypothetical person determined to hold a belief divergent from their own in moral, conventional, psychological, and metaphysi­cal belief domains. Clear trends of domain­specific developmental pat­terns emerged. Most notably, third­graders made significantly more negative judgments of holding divergent beliefs in general as compared to seventh­graders and college students. This developmental trend applied to all domains excluding morality, the least subjective category. Regardless of age, participants judged diversity of moral beliefs as well as acting on those beliefs to be unacceptable. Participants in all age groups also favored uniformity of belief within the moral and conven­tional domains and diversity of belief within the metaphysical domain.

Although these studies that focus on domain­specific differences examined individuals from ages eight to twenty­two, they prompt curiosity about domain dependency in younger populations. There has been a more concerted empirical focus on evaluating the manner in which children’s epistemological development changes as a func­tion of age as opposed to domain of belief. In our earlier study, we (Burr and Hofer, 2002) re­evaluated Mansfield and Clinchy’s (1985) study of epistemological awareness of three­, four­, seven­, and ten­year­olds in regard to domain dependence. Although unconsidered in Mansfield and Clinchy’s original discussion, we identified trends in the data as supportive of the domain dependence claim. Our re­analysis demonstrated that although preschoolers were generally identified as absolutists, elementary students progressed more quickly to a stage of multiplism on issues of taste than on matters concerning truth. The observed differences in rates of development according to domain offer initial support for examining younger children.

In an additional study, Wainryb et al. (2004) took initial steps to address this age gap, exploring how children in their early school years think about diversity of belief across different domains. The selection of domains differed from the subjectivity–objectivity continuum method of earlier studies (Kuhn et al., 2000; Kuhn and Weinstock, 2002), as Wainryb and colleagues employed a conceptual framework that divides the domains of “value” and “fact” into different categories. Impetus for this framework derives from evidence that children actively distinguish

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 233

among different types of value judgments concerning, for example, mat­ters of taste, morality, justice, or convention, and respond differently to them (Turiel, 1998; Mansfield and Clinchy, 2002). Children also appear to perceive and respond to different types of facts depending on whether the nature of the subjects they address is certain and irrefuta­ble or more open to personal interpretation (Carpendale and Chandler, 1996; Kalish, 1998; Kuhn et al., 2000; Kuhn and Weinstock, 2002).

Drawing on this research, the four domains chosen to contrast chil­dren’s thinking about diversity of beliefs were those of morality, taste, facts, and ambiguous facts (Wainryb et al., 2004). Groups of five­, seven­, and nine­year­old children were read vignettes about disagree­ing characters, only one of which, according to a baseline assessment, shared the actual view of the subject. Participants were assessed for judgments of relativism, tolerance, and the disagreeing person. Relativ­ism pertains to an assessment of whether participants believe that mul­tiple beliefs may be correct, or that only one belief is correct, whereas tolerance refers to whether children think it is acceptable or unaccept­able for people to endorse beliefs that conflict with their own.

Highly consistent realm­dependent developmental patterns were observed on judgments of relativism and tolerance. Five­year­olds made significantly fewer relative and tolerant judgments than seven­ and nine­year­olds and significantly more negative descriptions of disa­greeing characters. Domain also organized patterns of judgments: in the domains of moral and fact disagreements, participants, regardless of age, asserted that only one belief was right. However, significant age differences were discerned in judgments regarding the more subjective domains of ambiguous facts and taste. Only a third of five­year­olds judged multiple beliefs to be correct in these domains, in contrast to most of the seven­ and nine­year­olds. All participants were least toler­ant of divergent moral beliefs (Wainryb, 2004).

Most notably, however, Wainryb and colleagues (2004) excluded children younger than five years of age from this study under the assumption that theory of mind, and hence an understanding of the representational nature of beliefs, are not yet fully developed. How­ever, the exclusion of children younger than age five may ignore impor­tant precursors to theory of mind onset present in the epistemological understanding that precedes dualism.

Framework for our study

Our previous research suggests that between the ages of three and five children progress from the most primitive stage of egocentric subjectiv­ity through a transitional stage in which some epistemological awareness

Epistemological development in very young knowers234

is present, and finally, in a roughly simultaneous shift, achieve a solid theory of mind and enter the stage of absolutism (Burr and Hofer, 2002). Research on domain dependency in epistemological thought with chil­dren in this age range has been neglected, however. This empirical gap has been both motivated and justified by the assumption that pre­theory of mind children are unable to appreciate diversity of belief in any form, regardless of domain. Yet, there is a strong rationale to attempt research on domain dependency with this population. Clearly, children in the midst of developing a firm theory of mind are also grappling with fledg­ling forms of epistemological thought. Since children aged three to five display some epistemological awareness, and since the five­year­olds in Wainryb and colleagues’ (2004) study were not uniformly incapable of appreciating belief diversity, we determined that a study that examines domain dependency within this population was needed.

In this study, based on interviews developed and conducted by the first author (Wildenger, 2006), we focused on whether domain­dependent patterns of epistemological thought emerged prior to theory of mind mastery in three­ to five­year­old children. Contrary to the expecta­tion that pre­theory of mind children are entirely unable to recognize or accept that individuals can hold different beliefs in any of the four domains, we hypothesized that emerging patterns of domain depend­ence would be visible: some three­ to five­year­old children would be able to recognize that people can have different beliefs in subjec­tive domains (taste and ambiguous fact) more often than in objective domains (morality and fact), consistent with the way that Wainryb and colleagues’ (2004) five­year­old subjects considered belief diversity in these domains. We predicted that domain differentiation would emerge gradually, with some degree of acceptance of belief diversity preceding theory of mind development, given our earlier findings of subjectivity in three to five­year­olds (Burr and Hofer, 2002) and the emergence of dualism that accompanies theory of mind. (See Wildenger, 2006, for complete details of this second study.)

Method

Participants

The sample included thirty­seven participants (twenty females and sev­enteen males) in three age groups: three­year­olds (n = 10, M = 43.7 months), four­year­olds (n = 13, M =54.3 months), and five­year­olds (n = 14, M = 65.6 months). Participants, primarily middle class and Caucasian, were drawn from two private half­day preschool programs

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 235

in a small town in upstate New York, another private full­day preschool program in a small town in Vermont, and a full­day kindergarten class­room at a public elementary school in the same Vermont town.

Overview and procedure

Participants were interviewed by the first author in familiar classrooms at their schools. The interview included discussions of four epistemo­logical vignettes in various domains and two theory of mind tasks, and these sections were counterbalanced. Interviews were recorded on vid­eotape and subsequently transcribed.

Materials for the theory of mind tasks were replicated from Burr and Hofer’s (2002) study, and adapted from the procedure utilized by Lalonde and Chandler (1995). Two tasks involved unexpected change produced by a misplaced item, and two tasks involved a case of unex­pected contents. Domain dependency of epistemological thought was assessed through a replication of Wainryb et al.’s (2004) study. Partici­pants were told illustrative vignettes about two characters with conflict­ing beliefs in different domains. The presentation of the vignettes was accompanied by colorful 8.5 × 11 inch drawings to make it easier for the children to both understand and retain the information.2 Before the presentation of the vignettes, a baseline assessment was given to each child to determine his/her own belief about the topic and establish that the participant agreed with one of the characters and disagreed with the other (e.g., “Do you believe that rain is dry or wet?”). Consist­ent with Wainryb and colleague’s (2004) study, participants were told about four disagreements, each in a different domain, morality, taste, fact, and ambiguous fact. For example, “Sarah believes that it’s okay to hit and kick other children, and Sophie believes that it’s wrong to hit and kick other children.” Two versions of each disagreement were uti­lized, and children were randomly assigned to conditions. Gender was consistent within vignettes but varied between vignettes: both char­acters’ gender and presentation order of the four disagreements were counterbalanced across subjects with a Latin square design.

Scoring

Scoring for theory of mind followed Burr and Hofer’s (2002) method of summing the number of tasks successfully completed, resulting in a

2 The generosity of Cecilia Wainryb in providing her original drawings for use in this study is greatly appreciated by the authors.

Epistemological development in very young knowers236

score between 0 and 4. Those who scored 0–1 were classified as “low” theory of mind ability, whereas those who scored 2–4 were classified as “high” theory of mind ability.

Epistemological thought was scored consistent with Wainryb et al. (2004). Evidence of both relativity and tolerance in judgments of diverse beliefs is indicative of more advanced epistemological reasoning, according to the definition of multiplism (the ability to acknowledge the existence and merit of multiple viewpoints). Wainryb and colleagues’ relativism and tolerance interview questions were thus used to garner an assessment of epistemological thought. For each disagreement, the assessments were obtained through the following interview questions:

Relativism judgment. “Do you think that only one belief (what Sophie believes) is right, or do you think that both beliefs (what both Sophie and Sarah believe) are right?” (Participants who stated that only one belief is right were also asked, “Which one is right?”) “Why (is only one belief right/are both beliefs right)?”

Tolerance judgment. “Do you think it is ok for (disagreeing character) to believe (divergent belief) or do you think that it is not ok for him/her to believe that? Why is it ok/not ok for him/her to believe that?” (Wainryb et al., 2004, p. 691)

Judgments for tolerance and relativism were scored dichotomously, as either relativistic/tolerant (2) or not (1). An independent rater and the first author coded responses for all of the measures. Interrater reliabil­ity was determined via Cohen’s Kappa, with reliability of 0.94 for the theory of mind tasks, 1.00 for the relativity measures, and 0.97 for the tolerance measures.

Results

Analyses were conducted using two categorical variables, relativism and tolerance (both variables indicative of acceptance of diversity of belief ), and another dependent variable, theory of mind, dichotomized as low or high ability. The distribution of theory of mind scores in the sample was roughly even, with sixteen children scoring “low” and twenty­one children scoring “high” on the measure. Consistent with the existing literature (Wellman et al., 2001), a significant positive cor­relation between age and theory of mind ability was found, r(36) = 0.66, p < .01.

Epistemological thought

To determine whether epistemological thought varies with domain, age, or theory of mind ability among three­ to five­year­olds, we used

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 237

3 Percentages indicate the percent of all relative judgments in the study that were made by children in each age group.

repeated measures ANOVA, replicating the analysis strategy of Wain­ryb and colleagues (2004).

Relativism. Judgments of relativism concern whether only one or both of two conflicting beliefs are correct. To assess whether three­ to five­year­olds considered relativism differently according to domain and whether there were age differences within this population, a 4 × 3 factorial ANOVA was conducted by domain of disagreement and age, with domain as a repeated measure. The ANOVA yielded a significant main effect of domain, F(3,102) = 5.17, p < .05, η² = .13, a medium to large effect size (Cohen, 1988). Regardless of age, children made more relative judgments within the domain of ambiguous fact (48.6%, M = 1.50) than in either morality (18.9%, M = 1.22) or fact (21.6%, M = 1.24) domains.3 There was also a main effect of age, F(2,34) = 4.83, p < .05, η² = .22, a large effect size. Across domains, three­year­olds (52.5%, M = 1.53) made more relative judgments than five­year­olds (12.5%, M = 1.13). The age by domain interaction was not significant, F(6,102) = 1.20, p > .05.

To determine whether judgments of relativism were related to theory of mind ability for three­ to five­year­olds and whether this varied by domain, a second 4 x 2 factorial ANOVA was conducted by domain of disagreement and theory of mind ability (dichotomized as high or low), with domain as a repeated measure. The ANOVA yielded a main effect of domain, F(3,105) = 5.46, p < .05, η² = .14, a medium to large effect size such that, regardless of theory of mind ability, participants made more relative judgments in the ambiguous fact domain (48.6%, M = 1.50) than in the morality (18.9%, M = 1.21) or fact (21.6%, M = 1.24) domains. There was also a main effect of theory of mind ability, F(1,35) = 7.64, p < .01, η² = .18, a large effect size. Participants with low theory of mind ability (46.8%, M = 1.47) made more relative judgments than participants with high theory of mind ability (17.9%, M = 1.18), regardless of domain. The theory of mind by domain inter­action was not significant, F(3,105) = 1.15, p > .05.

Tolerance. Tolerance judgments concern whether it is acceptable or unacceptable to endorse a divergent belief. To assess whether three­ to five­year­olds considered tolerance differently according to domain and whether there were age differences within this population, an 4 x 3 factorial ANOVA was conducted by domain of disagreement and age, with domain as a repeated measure. The ANOVA yielded a main effect of domain, F(3,102) = 5.55, p < .05, η² = .14, a large effect size. Toler­ant judgments were significantly less frequent in the morality domain

Epistemological development in very young knowers238

(5.4%, M = 1.06) compared to both the fact (29.7%, M = 1.32) and ambiguous fact (27.0%, M = 1.27) domains, while tolerance judgments occurred more often in the fact domain (29.7%, M = 1.32) compared to the taste domain (13.5%, M = 1.15).4 Therefore, participants were the least tolerant within the morality domain and the most tolerant within the fact domain, regardless of age. No significant main effect of age was evident, F(2,34) = 1.78, p > .05, and the age by domain interaction also failed to reach significance, F(6,102) = 2.10, p > .05.

To determine whether judgments of tolerance were related to theory of mind ability for three­ to five­year­olds and whether this varied by domain, a second 4 x 2 factorial ANOVA was conducted by domain of disagreement and theory of mind ability (dichotomized as high or low), with domain as a repeated measure. The ANOVA yielded a main effect of domain, F(3,105) = 4.75, p < .05, η² = .12, a medium to large effect size. The pattern of tolerance judgments within specific domains mir­rored the results from the domain by age ANOVA, with the lowest level of tolerance in the morality domain (5.4%, M = 1.06), and the highest in the fact domain (29.7%, M = 1.31), regardless of theory of mind abil­ity. No significant main effect for theory of mind ability was discerned, F(1,35) = 1.42, p > .05, and the theory of mind by domain interaction also failed to reach significance, F(3,105) = .08, p > .05.

Discussion

Age, theory of mind, and the movement from subjectivity to objectivity

Several significant developmental differences emerged in this study. Both child age and theory of mind ability were related to patterns that emerged in the relative maturity of judgments of relativism and toler­ance among three­ to five­year­old participants. Since age and theory of mind ability were so highly correlated, the patterns produced by the two variables frequently mirrored one another in corresponding analy­ses (i.e., replacing age with theory of mind ability usually led to pre­cisely the same pattern of results).

Most importantly, three­year­olds and children with lower theory of mind ability made significantly more relative judgments than did five­year­olds and those with theory of mind ability. Thus we found that relativism declines with age between three and five years, and children

4 Percentages indicate the percent of all tolerant judgments in the study that were made by children in each age group.

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 239

appear to become increasingly more absolutist as they approach five, in tandem with their consolidation of an understanding of theory of mind. This finding is consistent with our earlier work (Burr and Hofer, 2002), in which we identified egocentric subjectivity and pre­dualism as precursors to absolutism, as well as with Wainryb et al.’s (2004) find­ings that five­year­olds are more absolutist than seven­ and nine­year­olds. Thus children may make highly relative and subjective judgments before theory of mind and then become rather rigid in their absolutism once they embrace the objectivity permitted by theory of mind. This appears to soften over time in the early elementary school years, as they begin to differentiate more among domains (Wainryb et al., 2004).

The five­year­olds in this study produced patterns of results that, although not completely consistent with their counterparts in Wainryb and colleagues’ study (2004), can be considered comparable. For example, although 100 percent of five­year­olds made non­relative judgments of morality and fact, 43 percent and 7 percent endorsed relative views in the ambiguous fact and taste domains, respectively. These participants often responded in typical absolutist terms, making very black­and­white distinctions between what they considered to be clearly correct and incorrect beliefs. Yet some variation in their accept­ance of belief diversity was present, as was expected. The four­year­olds in this study occupied a broad range of ability levels, as evidenced by widely varying theory of mind scores, findings that are consistent with the expectation that children in the middle age range would occupy a “transitional” epistemological phase.

Domain specificity

The domain, or realm of disagreement in which diversity of thought occurred, clearly organized three­ to five­year­old children’s judg­ments of both relativism (i.e., whether more than one belief can be cor­rect) and tolerance (i.e., whether it is acceptable to endorse a divergent belief ), consistent with our hypothesis that children within this age group would, in fact, differentiate between domains rather than uni­formly rejecting the idea of belief diversity.

We originally anticipated that children would be most accepting of belief diversity within subjective domains of taste and ambiguous fact, and least accepting within objective domains of morality and fact (Wainryb et al., 2004). Certain aspects of the results were consistent with this hypothesis. Participants made the least relative judgments within the morality domain and the most within the ambiguous fact domain. However, although we predicted that the domains of morality

Epistemological development in very young knowers240

and fact would emerge as a pair throughout the analyses, similarly to the domains of taste and ambiguous fact, this was often not the case. Instead, analyses revealed that participants were frequently more accept­ing of belief diversity within the fact domain than the taste domain.

Although there are undoubtedly multiple explanations for this trend, we will focus on some pertinent possibilities. The failure of three­ to five­year­old children to appreciate diversity within the taste domain, as expected, may reflect an immature understanding of conventions. Preschool­aged children tend to consider conventions as binding rules rather than social guidelines (Turiel, 1983). Very young children have difficultly understanding that conventions are relative rather than uni­versal (Smetana, 1981). The beliefs in the taste vignettes that “chocolate ice cream is yummy” and that “red flowers are pretty” are conventions that most children in this age group abide by. Therefore, based on this sense of what is highly normative among their peers, participants in this study often insisted that it was not possible or acceptable to believe, for example, that red flowers are ugly, because, as one child stated, “Flow­ers are not ugly, just beautiful. All flowers are beautiful, I think.” The possibility that conventions influenced participants’ taste judgments is a fertile area for future study. It would be interesting, for example, to alter the ice cream vignette so that the belief was less conventional by ques­tioning participants about strawberry ice cream rather than chocolate, or by making the choices more open­ended rather than dichotomous.

Another explanation for these unexpected domain­related trends focuses on fact disagreements. Although the factual statements used (i.e., “Rain is wet” and “pencils fall down”) are rather incontrovert­ible, participants tolerated conflicting beliefs in this domain to an unpredicted extent. It is a distinct possibility that the participants in this study interpreted the conflicting beliefs in the fact vignettes as pretense. There has been considerable interest in comparing children’s treatment of epistemological states such as belief to their treatment of fictional mental states such as pretense. Although children engaging in pretense do not intend to represent reality accurately, pretense and beliefs are both representational in nature (Woolley, 1995). Research generally suggests that prior to theory of mind development around the age of four or five, young children have difficulty distinguishing between belief and pretense (Kalish et al., 2000; Koenig, 2002; Wool­ley and Wellman, 1990, 1993). The main confusion exhibited by young children is uncertainty as to whether fictional mental states such as pretense represent reality (Woolley, 1995). In fact, Perner et al. (1994) deem young children’s fledgling understanding of these two concepts of pretense and belief “prelief.” Therefore, participants in this study may

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 241

have assumed that the disagreeing characters, by believing that “rain is dry” or “pencils fall up,” were simply pretending, an activity that they most likely considered acceptable, and would therefore tolerate. This potential confusion of belief with pretense, specifically within the fact domain, also requires future study. It would be helpful to conduct a pre­assessment of children’s understanding of belief when compared to pretending. It would also be interesting to isolate the fact domain and vary the language within the vignette, using either the word “pretend” or “believe” to discern whether any differences in acceptance of belief diversity emerged as a function of the epistemological state. Moreover, it might be fruitful to replace the word “believe” with “think” in these vignettes to assess whether this gives a different pattern of responses.

Limitations

Interviewing young children provides numerous challenges. During the interview process, it became clear that the epistemological vignettes were difficult for some of the three­year­olds. In retrospect, some ben­eficial revisions to the interview process might involve altering the lan­guage so it is less complex, shortening the length of the interview by reducing the number of questions, and supplementing vignettes with puppets, all with the aim of facilitating comprehension and retention. Many children required excessive prompting to both recall the beliefs of the characters and, in many cases, to recall even their own beliefs. This was a major difference from the Wainryb et al. (2004) study with older children, in which all of the participants were able to accurately recall and attribute the beliefs to the disagreeing characters in the vignettes. Understanding of the vignettes and the concept of belief itself may have posed too difficult a cognitive task for the epistemological stance of egocentric subjectivity. That is, because the youngest children in this study were unable to distinguish between knowledge and objective reality, assuming that everyone views the world in precisely the same manner, the concept of belief underlying the vignettes was essentially meaningless to them.

Often, it seemed that the youngest children were influenced by the recency effect (Aronson, 2004) within the relativism and tolerance measures. Asked to choose between two options, the youngest children frequently selected the latter choice, without otherwise indicating that they understood the question. This would contribute to the high level of relative, but not tolerant, judgments among three­year­olds and par­ticipants with low theory of mind ability. While the relativity measure was phrased “Do you believe that only one belief is right or that both

Epistemological development in very young knowers242

beliefs are right?” the tolerance measure happened to be phrased in the opposite order, “Do you think that it’s ok for Emily to believe that rain is dry or do you think it’s not ok for her to believe that?” Despite these concerns, the results do seem to indicate that there are meaningful dif­ferences in the ways that younger and older children consider belief diversity. The possibility exists that young children truly do understand the concept of relativism before they are able to appreciate tolerance.

Conclusions of this study

As a whole, these results indicate an emerging epistemological under­standing of children in this age group. It appears that a fully developed theory of mind is not a prerequisite for appreciating belief diversity, and that children become more objective and absolutist during the devel­opment of theory of mind. Perhaps the acceptance of belief diversity demonstrated by the younger participants lends credence to Chandler et al.’s (2002) recursion argument, one of multiple explanations for the identification of similar epistemological positions across the lifespan. According to these researchers, epistemological development may be conceptualized as a spiral­type progression rather than a linear path of development. In this model, the high level of acceptance of belief diversity among children with low theory of mind ability may signify an initial movement through an immature form of the later stage of multi­plism. Similar parallels have also been observed between, for example, populations of high school and college students.

Wainryb and colleagues concluded that “the generalized nega­tive view of diversity among 5­year­olds … consistent with the depic­tions of young children as objectivists and generally intolerant” (2004, p. 699) is an oversimplified underestimation of their ability to appre­ciate belief diversity. This study provides additional support for this claim. The five­year­olds in the current study were not uniformly unable to accept belief diversity, as the definition of absolutism would suggest. The present examination of epistemological thought in vari­ous domains with even younger participants, in the midst of theory of mind develop ment, has similarly elucidated the nuances of their emer­gent epistemological understanding. The assumption that these “tran­sitional” children are cognitively unable to consider belief diversity in any form is also an overgeneralization.

The epistemological thought of three­ to five­year­old children appears to parallel that of older children and adolescents as documented in previous studies within the morality and ambiguous fact domains (Kuhn et al., 2000; Mansfield and Clinchy, 2002; Wainryb et al., 2001,

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 243

2004). That is, children were notably more accepting of belief diversity when the discrepancies surrounded vague notions that were relatively open to interpretation. Children were significantly less accommodating of differences in opinion when disagreements involved ethical issues such as hurting another child or intentionally breaking toys. The unan­ticipated results within the taste and fact categories may indicate that children in this age group are still struggling to cognitively grasp cer­tain aspects of these domains.

Future research

Methodological considerations in assessing epistemology in young children

The quest to understand the content and character of young children’s thinking is not a new one. Formally beginning with Piaget’s (1950/1970) groundbreaking research on cognitive development, researchers have spent decades inventing creative strategies and methods for peering inside the “black box” of young children’s minds. Today, those inter­ested in establishing the early roots of epistemological development face many of the same challenges as our scientific predecessors. Creating tasks and methods that will provide meaningful information about the content of young children’s understanding of knowledge and knowing will be difficult. However, some of the lessons from broader studies of children’s cognition may serve as a valuable starting place for these investigations.

Basic cognitive functions. The assessment of children’s epistemological development must be predicated on an understanding of the changes in basic cognitive functioning that occur during the early years, as these changes strongly influence a child’s ability to participate in structured tasks and procedures. Two examples are the development of executive function and the growth of working memory. Emerging research on the executive function highlights that a child’s ability to focus, main­tain, and shift attention undergoes substantial enhancement during the preschool years (e.g., Mutter et al., 2006). There is also strong evidence that a child’s working memory capacity expands rapidly during the early school years, allowing for increased digit span, complex or “working” memory span, and visuo­spatial recall (Gathercole et al., 2004).

As a result of these developmental changes, there are several steps that researchers can take to ensure that children can best demonstrate their knowledge during data collection sessions. First, we recom­mend that data collection be conducted in the context of individual,

Epistemological development in very young knowers244

one­on­one interviews. This allows the researcher to monitor children’s attention span and understanding of the tasks. Second, we recommend that these interview sessions be held to a maximum of fifteen minutes for preschoolers and twenty minutes for young elementary students. In our experience, children are typically quite willing to participate in multiple sessions, and they are better able to maintain attention over a shorter period of time. Third, we also recommend that researchers work to make the interviews as interactive as possible. The false­belief tasks commonly used to assess theory of mind (e.g., Wellman et al., 2001) and the tasks developed to examine children’s understanding of the sources of their knowledge (e.g., O’Neill et al., 1992; O’Neill and Gopnik, 1991) are good examples of methods that draw children in and assist them in remembering the scenario being described. For exam­ple, in the O’Neill et al. (1992) tasks, children are asked to reach into a tunnel and indicate whether or not they can determine the identity of a hidden object. Children are then asked to infer the knowledge of a puppet that is using the same sensory modality. In our lab, we have similarly attempted to engage children through the use of props, pup­pets, and cartoons. In our experience, the more interactive our meth­ods, the easier time children seem to have remaining on task and fully participating. Researchers are encouraged to continue this practice in future studies.

Cognitive or developmental delays. A growing body of research sup­ports the supposition that children and adults who have been diagnosed with autism frequently have difficulty passing theory of mind tasks (e.g., Martin, 1999). As a result, future examinations of epistemologi­cal development should include a screening process where children who have been diagnosed with autism are either excluded from the study or analyzed separately. Similarly, researchers should consider excluding children who have been diagnosed with other developmental delays, given potential decreased working memory ability or attention span, as discussed above. Based on studies in our lab, for example, we have found that the theory of mind ability of five­year­olds who are in pre­school differs from the theory of mind ability of five­year­olds who are in kindergarten. Subtle developmental delays or individual differences in school readiness are likely related to this finding, and ignoring these differences can make the data difficult to interpret. Researchers should carefully consider the source of their participants and use screening measures in an effort to gain a representative sample with few potential confounds.

Verbal ability. The development of language is a monumental task that dramatically influences children’s understanding of the world

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 245

around them. Given the private nature of children’s cognitive beliefs, verbal language is necessary to convey their ideas and understandings of knowledge, learning, and beliefs. This trend is mirrored in much of the developmental literature. Much of what we know about the interior lives of young children relies on verbal self­report (e.g., preoperational thinking, Piaget, 1950/1970).

Future studies of children’s epistemological development should include measures of verbal ability, including both receptive and produc­tive vocabulary. Then, during statistical analyses, the influence (i.e., proportion of variance) of verbal ability can be partitioned away from the primary relations of interest. For example, in one study, preschool­ers’ verbal ability (as measured by a series of language tasks includ­ing both syntax and semantics) accounted for a dramatic 36 percent of the variance in their concurrent performance on theory of mind tasks (Slade and Ruffman, 2005). If researchers use tasks that require verbal responses and fail to measure children’s verbal ability, there is always the possibility of ambiguous findings. Do the children who fail to artic­ulate the “correct” answers to the tasks genuinely hold a less developed view, or are they simply unable to voice their understanding, given their limited verbal capabilities? For example, in our second study, many children required excessive prompting by the researcher to both recall the beliefs of the characters and, in many cases, to recall even their own beliefs. Several of the youngest children also responded to ques­tions with “I don’t know.” In these cases, it was difficult to separate the children’s actual knowledge and understanding from their ability to articulate those thoughts and beliefs. It would be fruitful for future studies to better examine the relation of verbal ability to epistemologi­cal understanding.

Typically, children’s receptive vocabulary exceeds their expressive vocabulary (Nurss and Day, 1971; Turner and Rommetveit, 1967). As a result, researchers should make an effort to structure their tasks so that children can provide a non­verbal response to the experimenter’s questions. For example, in our lab we have structured tasks to include yes/no responses to prompts (allowing children to nod their heads in response). We’ve also used tasks where children can indicate their cur­rent or prior knowledge by pointing to a specific location, using props (Burr and Hofer, 2002). Incorporating non­verbal response options may help to alleviate some of the potential confounds noted above and should allow for more reliable assessment of younger participants.

Researchers may wish to move beyond verbal tasks to find non­ verbal methods of assessing children’s epistemological understanding. It is possible to envision a task somewhat similar to DeLoache’s classic

Epistemological development in very young knowers246

shrinking room extended to assess epistemological understanding (e.g., Deloache et al., 1997). In the shrinking room task, children are intro­duced to a miniature­scale model of a room and are shown where an object is hidden in the model. The children are then led into a full­size room that exactly matches the model. The children are then observed to see if they use the miniature room as a representation of the larger room and subsequently look for the hidden object in the same place it was located in the model. Researchers should consider how the various domains of epistemological understanding might be assessed observa­tionally through contrived situations.

Future directions

As our current understanding of epistemological development during early childhood is in its fledgling state, there are numerous important directions for future work. However, moving forward from studies in our own lab and those of others (e.g., Mansfield and Clinchy, 1985, 2002; Wainryb et al., 2004), there are some areas that seem particularly deserving of attention.

Perhaps the most important goal for researchers studying epistemic beliefs in young knowers is to find ways to pull apart the separate dimen­sions of epistemological understanding. A complete picture of early epistemological development will not emerge until we have evaluated children’s: (1) understanding of sources of knowledge; (2) valuation of the importance and reasons for justifying knowledge; (3) views of the certainty of knowledge; (4) understanding of the simplicity of knowl­edge; and (5) perception of whether or not these dimensions change by domain (for review, see Hofer and Pintrich, 1997).

The meaning of “domain” needs considerable clarification (Hofer, 2006). The differences in how the term has been operationalized in various developmental studies has been problematic, and yet under­standable at this early stage of the research. Researchers would benefit, however, from a more standardized definition of the particular dimen­sions under consideration.

Although cross­sectional studies can provide us with important insight into age­related trends, the only way to truly assess development is through the use of longitudinal studies. The sole means to disentan­gle the numerous claims about recursive development ( Chandler et al., 2002), domain­specific development (Kuhn and Weinstock, 2002), and the nature of early beliefs (Burr and Hofer, 2002) will be to employ lon­gitudinal methods. Researchers would also be advised to consider the

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 247

possibility that the same core beliefs may manifest themselves differ­ently at different ages (i.e., heterotypic continuity; Rutter et al., 2006). A three­year­old and a college student who both believe that knowledge about moral beliefs is uncertain, worthy of justification, and subjective will likely communicate and describe these beliefs in very different, age­appropriate ways.

In our lab, we have worked to connect our research on early epis­temological development to other existing studies of early cognitive milestones (e.g., theory of mind; Wellman et al., 2001). We believe that future research will benefit from continuing these connections. For example, “basic” research on theory of mind has progressed beyond the preschool years to examine second­order false beliefs (i.e., the under­standing that someone can hold a false belief about another’s beliefs) in elementary­aged children (Astington et al., 2002). Much as the “initial” mastery of false­belief tasks has been proposed as a meaningful land­mark in children’s epistemological development (e.g., Burr and Hofer, 2002), it seems likely that this second­order mastery would similarly have important epistemological relevance. By continuing to connect literature from different “areas” of cognitive development, researchers can continue to create a more thorough understanding of young chil­dren’s folk epistemology.

In summary, future research should continue to address how epis­temological thought relates to the ways that children learn and to the views that they hold of themselves as learners. Researchers also need to continue to address the ways that both epistemological and theory of mind development relate to constructs such as social skills and moral development that shape interpersonal interactions among children. It is also critical for studies examining very young children to continue to address the relations among epistemological development, theory of mind, and other cognitive accomplishments such as egocentrism, understanding of conventions, and pretense. An overarching aim of future domain­dependency research in epistemological development should be to clarify discrepancies among proposed epistemological positions across the lifespan.

Finally, studies of epistemological development in high­school and college­aged participants have, in large part, been valuable because of their links to educational experiences. Similarly, there is good reason to believe that epistemic beliefs may take on unique importance as children move into kindergarten and enter more formal learning environments. Researchers would be well­advised to examine the ways in which epis­temic understanding changes during the transition to kindergarten.

Epistemological development in very young knowers248

Educational implications

Most young children are enrolled in play­based early education pro­grams (pre­kindergarten) and are not yet involved with formalized aca­demic instruction. Accordingly, epistemological research with this age group perhaps has fewer direct consequences for academic motivation and performance compared to research with older children and adults. Yet research on epistemological development and theory of mind has implications for early childhood education. Our findings suggest that early childhood educators need to be aware of epistemological consid­erations as they teach their students. For example, because several chil­dren self­corrected their initially wrong responses when asked to justify their answers in the Burr and Hofer (2002) study, we recommend that adults assist children in making correct judgments by asking preschool­ers to justify their thinking.

Additionally, a preschooler’s first answer may differ from a more carefully considered response. These issues are especially important in assessment situations. While preschoolers are not a group frequently assessed via standardized tests and examinations, they are presented with many questions on a daily basis. For example, teachers might request that children explain the content of creative projects or draw­ings or describe strategies for solving simple problems. Teachers could also frequently encourage preschool students to answer questions con­cerning story content while they are reading aloud in order to engage them, encourage language use, and promote understanding of impor­tant concepts. When questions involve interpretations of character’s feelings and thoughts (e.g., “How do you think the little girl felt after she lost her teddy bear?”) and thus draw on theory of mind abilities, it would seem particularly important for teachers to encourage children to justify their thinking while remaining sensitive to and accepting of their level of epistemic understanding. By encouraging children to think about issues such as perspective­taking through fictional stories and pretend play, teachers can challenge children to consider viewpoints that are increasingly sophisticated by presenting new information in ways that reflect and are sensitive to their young students’ epistemo­logical development.

The kindergarten transition. The transition from preschool to kinder­garten represents a developmental landmark for children. As children move from typically play­oriented preschool programs to more struc­tured kindergarten environments, they encounter substantial differ­ences in many important educational variables that may potentially influence epistemological development. Formal school entry places

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 249

children in the official role of student, thus they are faced with more structured and challenging academic demands. Unlike preschool, the kindergarten curriculum in the US is often grounded in “formalized instruction” with the intent to increase children’s skill levels (Rimm­Kaufman et al., 2000). Thus, children may face formal evaluation of their academic skills, such as reading and writing, for the first time in kindergarten (Perry Weinstein, 1998). The student–teacher relation­ship may also change in nature, as kindergarten teachers are likely more concerned with academic progress than preschool teachers and generally must instruct a larger number of students (Rimm­Kaufman and Pianta, 2000). Due to the greater proportion of time allocated to academic activities, there is a simultaneous decrease in unstructured, play­based activity in kindergarten. Finally, kindergarten environments in the US are generally characterized by more rigid systems of rules and boundaries to assist students in learning the behavioral expectations for formal educational environments.

The transition to kindergarten represents a rough approximation of the advent of both established theory of mind understanding and the shift to an absolutist epistemic stance for many children. Thus, this event represents a unique juncture in the epistemological development of young children, with the potential for epistemic beliefs to take on special meaning. Children, now in the formal role of student, may first conceive of themselves as true learners in kindergarten. They may also begin to make meaningful distinctions among knowledge in different academic domains. Children are also likely to begin to conceptual­ize knowledge differently in this new educational context. Faced with formal achievement expectations, kindergarten students may begin to evaluate knowledge as either entirely correct or incorrect. Interestingly, this type of rather inflexible distinction is consistent with the epistemic stance of absolutism. Thus, the qualitative shifts that characterize the transition to kindergarten may reinforce the typical developmental pro­gression of epistemic beliefs.

Sensitivity to children’s epistemological development during this period has the potential to inform developmentally appropriate kin­dergarten instruction. For example, because preschool children learn about perspective­taking through imaginative play (e.g., Connolly and Doyle, 1984), it may be beneficial for teachers to continue to incor­porate play as part of a balanced kindergarten curriculum in addition to a focus on explicit academic instruction. In addition, although it may be effective to encourage older children to approach academic tasks by creatively employing a variety of problem­solving strategies, this instructional tactic may be confusing to kindergarten students, as

Epistemological development in very young knowers250

epistemic absolutists who are thoroughly convinced of the existence of a single correct answer.

Contemporary academic expectations for kindergarten students are increasing rapidly. This trend is likely a partial reflection of both the prevalence of full­day kindergarten programs (West et al., 2000) and the national educational movement toward increased accountability of schools. The No Child Left Behind Act of 2001 requires schools to improve student performance to nationwide benchmarks or risk probation (US Department of Education, 2002). Thus, schools feel pressure to begin preparing even their youngest students to achieve on high­stakes tests. The traditional first­grade curriculum has been pro­gressively “pushed down” to the kindergarten level as a result. Thus, kindergarten has become increasingly more academic as instruction continues to increase in speed and intensity and children are expected to master more skills at an earlier age (National Education Goals Panel, 1998). For example, a recent study (Booher­Jennings, 2005) noted that many kindergarten teachers have opted to reduce recess time dramati­cally in order to create more time for academic instruction. Educators also feel pressure to rely heavily on traditional, teacher­centered, paper­and­pencil instructional approaches with their kindergarten students in order to meet these standards (Booher­Jennings, 2005). This reduces kindergarten teachers’ f lexibility to use developmentally appropriate practices (e.g., play, integrated instruction) in their class­rooms (Goldstein, 2007).

This shift has major implications for the objectives of early childhood education as well as for the personal epistemology of young students beginning school. If the goals for kindergarten students are concep­tualized as heavily weighted toward academics, there will be less time devoted to encouraging social development through unstructured play. This de­emphasis on play threatens to negatively impact children’s epistemological development, as many contend that children’s engage­ment in shared pretend play promotes both social competence and an understanding of other people’s feelings and beliefs that underlies theory of mind mastery (e.g., Youngblade and Dunn, 1995; Dockett, 1998). Furthermore, traditional, didactic instructional techniques as opposed to more unstructured, exploratory styles of learning may be less sensitive to kindergarten students’ emergent epistemic beliefs about themselves as learners. Additionally, this trend toward accountability has filtered down to preschools, where teachers also feel an increased need to respond to district accountability requirements in structuring their curriculum to include more academic content (e.g., Desimone et al., 2004). Addressing the tension between accountability pressures

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 251

and remaining sensitive to young children’s epistemic levels poses a sig­nificant challenge for early childhood educators.

Social development. Research on early epistemological development is particularly rich with implications for social development in early educational settings. The recent research in our lab concerning domain dependency in epistemological thought suggests that there are impor­tant social ramifications for teachers and young students. As Wainryb and colleagues (2004) suggest, a developmental task in the epistemolog­ical thought of children is “Rather than learning to process all instances of diversity through a particular mode of thinking (such as a tolerant attitude) … to learn to recognize the features that distinguish among different types of differences” (2004, p. 702). Results from the current study suggest that children may be beginning to discriminate among domains of knowledge as early as the preschool years. Therefore, it may be problematic for educators and parents to attempt to instill a generally tolerant attitude in children of this age, as well. Contemporary educa­tional trends such as the inclusion movement and the increasing diver­sity of student populations have resulted in a heightened emphasis on acceptance and tolerance of individual differences. Yet it is important to realize that this attitude may become over­generalized by young chil­dren. Our research suggests that it may be more appropriate to teach even very young children to discriminate among types of conflicts in deciding whether multiple perspectives are feasible or desirable, accord­ing to Wainryb’s recommendations (2004). This may give early educa­tors added insight into effective strategies for resolving differences of opinion among their students.

In a broader sense, the cognitive theory of mind and epistemological accomplishments of young children are explicitly linked to their social adjustment skills in early educational settings. For example, research has demonstrated that children’s understanding of others’ mental states is positively correlated with the quality of both their social interactions and their peer relationships in school (e.g., Dunn and Cutting, 1999; Dunn et al., 2000). Child social skills are widely considered to be as critical as, if not more critical than, academic skills such as counting and knowing the alphabet for success in early educational settings (e.g., McIntyre et al., 2006; Rimm­Kaufman et al., 2000). In fact, research indicates that kindergarten teachers consider social skills such as shar­ing, taking turns, being sensitive to the feelings of others, and contrib­uting appropriately in group contexts to be “survival skills” necessary for success in the transition to formal schooling (e.g., Dockett and Perry, 2001; Fowler et al., 1991). These types of social skills facilitate the development of positive peer relations which, in turn, predict positive

Epistemological development in very young knowers252

long­term educational outcomes, both academic and social (e.g., Ladd, 1990; Walker et al., 1995). Thus, the cognitive abilities associated with theory of mind and epistemological development may play a significant role in socio­behavioral skills associated with a successful transition to kindergarten for young children.

It is important for early educators to be aware that epistemological development may emerge differently for children with developmental disabilities. In particular, the relationship between autism and theory of mind has received considerable attention. Children with autism have been found to have considerable difficulty understanding the concept of false beliefs (e.g., Baron­Cohen et al., 1985). Many espouse the theo­ry­of­mind hypothesis of autism and contend that children with autism spectrum disorders experience a deficit in mental state understand­ing, or theory of mind ability, which may help to explain some of the core deficits of the disorder, including impairments in social and com­munication development (e.g., Begeer et al., 2003; Serra et al., 2002; Tager­Flusberg, 2001). Today, more children than ever before are being diagnosed with autism in early childhood. Current prevalence rates indicate that approximately 1 in 150 children have an autism spec­trum disorder (US Department of Health and Human Services, 2007). Children with poor social skills have been empirically documented to experience a more negative transition to kindergarten (McIntyre et al., 2006). Thus, children with autism spectrum disorders may be at a higher risk for a difficult transition due to their theory of mind deficits and impairments in reciprocal social interaction. Given its prevalence, early childhood educators are likely to encounter children with autism in their classrooms. It is imperative that both preschool and kindergarten teachers maintain a particularly sensitive awareness of the potentially stark epistemic differences between children with autism and their typically developing peers and encourage their social develop­ment in appropriate ways.

To equip all of their students with skills for success in kindergarten, teachers can encourage perspective­taking and empathetic, reciprocal social interactions in preschool. In order to help prepare their students for a positive transition to formal schooling and long­term educational success, it is therefore essential that early childhood educators are sen­sitive to the epistemological development of very young knowers.

R EF ER ENCES

Aronson, E. (2004). The social animal (9th edn.). New York: Worth.Astington, J. W., Pelletier, J., and Homer, B. (2002). Theory of mind and epis­

temological development: The relation between children’s second­order

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 253

false­belief understanding and their ability to reason about evidence. New Ideas in Psychology, 20, 131–44.

Baron­Cohen, S., Leslie, A. M., and Frith, U. (1985). Does the autistic child have a “theory of mind”? Cognition, 21, 37–46.

Baxter Magolda, M. B. (1992). Knowing and reasoning in college: Gender-related patterns in students’ intellectual development. San Francisco: Jossey­Bass.

Begeer, S., Rieffe, C., Terwogt, M., and Stockmann, L. (2003). Theory of mind­based action in children from the autism spectrum. Journal of Autism and Developmental Disorders, 33(5), 479–87.

Belenky, M. F., Clinchy, B. M., Goldberger, N. R., and Tarule, J. M. (1986). Women’s ways of knowing: The development of self, voice and mind. New York: Basic Books.

Booher­Jennings, J. (2005). Below the bubble: “Educational triage” and the Texas accountability system. American Educational Research Journal, 42(2), 231–68.

Bredekamp, S. and Copple, C. (Eds.) (1997). Developmentally appropriate prac-tice in early childhood programs. Washington, DC: National Association for the Education of Young Children.

Brooks­Gunn, J., Rouse, C. E., and McLanahan, S. (2007). Racial and eth­nic gaps in school readiness. In. R. C. Pianta, M. J. Cox, and K. L. Snow (Eds.), School readiness and the transition to kindergarten in the era of account-ability (283–306). Baltimore, MD: Brookes.

Burr, J. E., and Hofer, B. K. (2002). Personal epistemology and theory of mind: Deciphering young children’s beliefs about knowledge and know­ing. New Ideas in Psychology, 20, 199–224.

Carpendale, J. I. and Chandler, M. J. (1996). On the distinction between false belief understanding and subscribing to an interpretive theory of mind. Child Development, 67, 1686–706.

Chandler, M. J. and Carpendale, J. I. (1998). Inching toward a mature theory of mind. In M. D. Ferrari and R. J. Sternberg (Eds.), Self-awareness: Its nature and development (148–90). New York: Guilford.

Chandler, M. J., Hallett, D., and Sokol, B. W. (2002). Competing claims about competing knowledge claims. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (145–68). Mahwah, NJ: Lawrence Erlbaum Associates.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd edn.). Hillsdale, NJ: Lawrence Erlbaum Associates.

Connolly, J. and Doyle, A. (1984). Relation of social fantasy play to social com­petence in preschoolers. Developmental Psychology, 20, 797–806.

DeCicca, P. (2007). Does full­day kindergarten matter? Evidence from the first two years of schooling. Economics of Education Review, 26, 67–82.

DeLoache, J. S., Miller, K., and Rosengren, K. S. (1997). The credible shrink­ing room: Very young children’s performance with symbolic and non­symbolic relations. Psychological Science, 8(4), 308–13.

Desimone, L., Payne, B., Fedoravicius, N., Henrich, C., and Finn­Stevenson, M. (2004). Comprehensive school reform: An implementation study of preschool programs in elementary schools. The Elementary School Journal, 104(5), 369–89.

Epistemological development in very young knowers254

Dockett, S. (1998). Constructing understandings through play in the early years. International Journal of Early Years Education, 6(1), 105–16.

Dockett, S. and Perry, B. (2001). Starting school: Effective transitions. Early Childhood Research and Practice, 3(2). Retrieved December 14, 2006, from http://ecrp.uiuc.edu/v3n2/dockett.html.

Dunn, J. and Cutting, A. L. (1999). Understanding others, and individual dif­ferences in friendship interactions in young children. Social Development, 8(2), 201–19.

Dunn, J., Cutting, A. L., and Demetriou, H. (2000). Moral sensibility, under­standing others, and children’s friendship interactions in the preschool period. British Journal of Developmental Psychology, 18(2), 159–77.

Flavell, J. H. (1999). Cognitive development: Children’s knowledge about the mind. Annual Review of Psychology, 50, 21–45.

Fowler, S., Schwartz, I., and Atwater, J. (1991). Perspectives on the transi­tion from preschool to kindergarten for children with disabilities and their families. Exceptional Children, 58(2), 136–45.

Gathercole, S. E., Pickering, S. J., Ambridge, B., and Wearing, H. (2004). The structure of working memory from 4 to 15 years of age. Developmental Psychology, 40(2), 177–90.

Goldstein, L. (2007). Beyond the DAP versus standards dilemma: Examining the unforgiving complexity of kindergarten teaching in the United States. Early Childhood Research Quarterly, 22(1), 39–54.

Graue, E. (2001). What’s going on in the children’s garden? Kindergarten today. Research in review. Young Children, 56(3), 67–73.

Hatch, J. A. (2002). Accountability shovedown: Resisting the standards move­ment in early childhood education. Phi Delta Kappan, 83(6), 457–62.

Hofer, B. K. (2001). Personal epistemology research: Implications for learning and teaching. Educational Psychology Review, 13(4), 353–83.

(2006). Beliefs about knowledge and knowing: Integrating domain specifi­city and domain generality. Educational Psychology Review, 18, 67–76.

Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learn­ing. Review of Educational Research, 67(1), 88–140.

Kalish, C. (1998). Natural and artifactual kinds: Are children realists or rela­tivists about categories? Developmental Psychology, 34(2), 376–91.

Kalish, C., Weissman, M., and Bernstein, D. (2000). Taking decisions seri­ously: Young children’s understanding of conventional truth. Child Devel-opment, 71(5), 1289.

King, P. M. and Kitchener, K. S. (1994). Developing reflective judgment: Under-standing and promoting intellectual growth and critical thinking in adolescents and adults. San Francisco: Jossey­Bass.

Kitchener, R. F. (2002). Folk epistemology: An introduction. New Ideas in Psy-chology, 20, 89–105.

Koenig, M. A. (2002). Children’s understanding of belief as a normative con­cept. New Ideas in Psychology, 20, 107–30.

Kuhn, D. (1991). The skills of argument. Cambridge: Cambridge University Press.

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 255

(2000). Theory of mind, metacognition, and reasoning: A life­span perspec­tive. In P. Mitchell and K. J. Rigg (Eds.), Children’s reasoning and the mind (301–26). Hove, England: Psychology Press.

Kuhn, D., Cheney, R., and Weinstock, M. (2000). The development of episte­mological understanding. Cognitive Development, 15, 309–28.

Kuhn, D. and Weinstock, M. (2002). What is epistemological thinking and why does it matter? In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

Ladd, G. (1990). Having friends, keeping friends, making friends, and being liked by peers in the classroom: Predictors of children’s early school adjust­ment? Child Development, 61(4), 1081–100.

Lalonde, C. E. and Chandler, M. J. (1995). False belief understanding goes to school: On the social­emotional consequences of coming early or late to a first theory of mind. Cognition and Emotion, 9(2/3), 167–85.

Lee, K. and Homer, B. (1999). Children as folk psychologists: The developing understanding of the mind. In A. Slater et al. (Eds.), The Blackwell reader in development psychology (228–52). Malden, MA: Blackwell Publishers.

Mansfield, A. F. and Clinchy, B. M. (1985). Early growth of multiplism in the child. Paper presented at the fifteenth annual symposium of the Jean Piaget Society, Philadelphia, PA.

(2002). Toward the integration of objectivity and subjectivity: Epistemologi­cal development from 10 to 16. New Ideas in Psychology, 20, 225–62.

Martin, N. (1999). Autism and theory of mind. Psychologist, 12(10), 516–17.Mason, L., Boldrin, A., and Zurlo, G. (2006). Epistemological understand­

ing in different judgment domains: Relationships with gender, grade level, and curriculum. International Journal of Educational Research, 45, 43–56.

McIntyre, L. L., Blacher, J., and Baker, B. L. (2006). The transition to school: Adaptation in young children with and without intellectual dis­ability. Journal of Intellectual Disability Research, 50(5), 349–61.

Muis, K. R., Bendixen, L. D., and Haerle, F. C. (2006). Domain­generality and domain­specificity in personal epistemological research: Philosophi­cal and empirical reflections in the development of a theoretical frame­work. Educational Psychology Review, 18, 3–54.

Mutter, B., Welsh, M., Alcorn, M. B. (2006). Theory of mind and executive function: Working­memory capacity and inhibitory control as predic­tors of false­belief task performance. Perceptual and Motor Skills, 102(3), 819–35.

National Education Goals Panel (1998). Ready schools. Washington, DC: National Education Goals Panel.

NICHD Early Child Care Research Network (2005). Characteristics and quality of child care for toddlers and preschoolers. In NICHD Early Child Care Research Network (Ed.), Child care and child development: Results from the NICHD study of early child care and youth development (91–102). Guilford Press: New York.

Epistemological development in very young knowers256

Nurss, J. R. and Day, D. E. (1971). Imitation, comprehension, and production of grammatical structures. Journal of Verbal Learning and Verbal Behavior, 10, 68–74.

O’Neill, D. K., Astington, J. W., and Flavell, J. H. (1992). Young children’s understanding of the role that sensory experiences play in knowledge acquisition. Child Development, 63, 474–90.

O’Neill, D. K. and Gopnik, A. (1991). Young children’s ability to identify the sources of their beliefs. Developmental Psychology, 27(3), 390–7.

Perner, J., Baker, S., and Hutton, D. (1994). Prelief: The conceptual origins of belief and pretence. In C. Lewis and P. Mitchell (Eds.), Children’s early understanding of mind: Origins and development (261–86). Hillsdale, NJ: Lawrence Erlbaum Associates.

Perry, W. G. (1970). Forms of intellectual and ethical development in the college years: A scheme. New York: Holt, Rinehart and Winston.

Perry, K. E. and Weinstein, R. S. (1998). The social context of early schooling and children’s school adjustment. Educational Psychologist, 33(4), 177–94.

Piaget, J. (1929/1951). The child’s conception of the world (J. Tomlinson and A. Tomlinson, trans.). London: Routledge and Kegan Paul.

(1950/1970). Genetic Epistemology (E. Duckworth, trans.). New York: Colum­bia University Press.

Rimm­Kaufman, S. and Pianta, R. (2000). An ecological perspective on the transition to kindergarten: A theoretical framework to guide empirical research. Journal of Applied Developmental Psychology, 21(5), 491–511.

Rimm­Kaufman, S., Pianta, R., and Cox, M. (2000). Teachers’ judgments of problems in the transition to kindergarten. Early Childhood Research Quar-terly, 15(2), 147–66.

Rutter, M., Kim­Cohen, J., and Maughan, B. (2006). Continuities and discon­tinuities in psychopathology between childhood and adult life. Journal of Child Psychology and Psychiatry, 47(3–4), 276–95.

Serra, M., Loth, F., van Geert, P., Hurkens, E., and Minderaa, R. (2002). Theory of mind in children with lesser variants of autism: A longitudinal study. Journal of Child Psychology and Psychiatry, 43(7), 885–900.

Slade, L. and Ruffman, T. (2005). How language does (and does not) relate to theory of mind: A longitudinal study of syntax, semantics, working memory, and false belief. British Journal of Developmental Psychology, 23, 117–41.

Smetana, J. (1981). Preschool children’s conceptions of moral and social rules. Child Development, 52(4), 1333–6.

Tager­Flusberg, H. (2001). Understanding the language and communicative impairments in autism. In L. Glidden (Ed.), International review of research in mental retardation: Autism, Vol. 23 (185–205). San Diego: Academic Press.

Turiel, E. (1983). The development of social knowledge: Morality and convention. Cambridge: Cambridge University Press.

(1998). The development of morality. In W. Damon (Series Ed.) and N. Eisenberg (Vol. Ed.), Handbook of child psychology, Vol. 3: Social, emo-tional, and personality development (5th edn., 863–932). New York: Wiley.

Leah K. Wildenger, Barbara K. Hofer, and Jean E. Burr 257

Turner, E.A. and Rommetveit, R. (1967). The acquisition of sentence voice and reversibility. Child Development, 38, 649–60.

US Department of Health, and Human Services (2002). No Child Left Behind (NCLB) Act of 2001 (NCLB). Public Law 1–7–110. Retrieved October 4, 2007, from www.ed.gov/policy/elsec/leg/esea02/index.html.

(2007). Autism information center frequently asked questions: Prevalence. Cent­ers for Disease Control and Prevention. Retrieved October 6, 2007 from www.cdc.gov/ncbddd/autism/faq_prevalence.htm#whatisprevalence.

Wainryb, C., Shaw, L. A., Langley, M., Cottam, K., and Lewis, R. (2004). Children’s thinking about diversity of belief and in the early school years: Judgments of relativism, tolerance, and disagreeing persons. Child Development, 75, 687–703.

Wainryb, C., Shaw, L. A., Laupa, M., and Smith, K. (2001). Children’s, ado­lescents’, and young adults’ thinking about different types of disagree­ments. Developmental Psychology, 37, 373–86.

Walker, H., Colvin, G., and Ramsey, E. (1995). Antisocial behavior in school: Strategies and best practices. Pacific Grove, CA: Brooks/Cole Pub­lishing Co.

Wellman, H. M., Cross, D., and Watson, J. (2001). Meta­analysis of theory­of­mind development: The truth about false belief. Child Development, 72(3), 655–84.

West, J., Denton, K., and Germino­Hausken, E. (2000). America’s kinder-gartners. US Department of Education, National Center for Education Statistics (NCES 2000–070). Washington, DC: US Government Printing Office.

Wildenger, L. (2006). Theory of mind and domain specificity of epistemological thought in 3- to 5-year-old children. Unpublished thesis, Middlebury Col­lege, Middlebury, VT.

Woolley, J. D. (1995). The fictional mind: Young children’s understanding of imagination, pretense, and dreams. Developmental Review, 15, 172–211.

Woolley, J. D. and Wellman, H. M. (1990). Young children’s understandings of realities, nonrealities, and appearances. Child Development, 61, 946–61.

(1993). Origin and truth: Young children’s understandings of imaginary mental representations. Child Development, 64, 1–17.

Youngblade, L. and Dunn, J. (1995). Individual differences in young chil­dren’s pretend play with mother and sibling: Links to relationships and understanding of other people’s feelings and beliefs. Child Development, 66, 1472–92.

Zigler, E. and Finn­Stevenson, M. (2007). From research to policy and prac­tice: The school of the 21st century. American Journal of Orthopsychiatry, 77(2), 175–81.

258

9 Beliefs about knowledge and revision of knowledge: on the importance of epistemic beliefs for intentional conceptual change in elementary and middle school students

Lucia Mason University of Padua, Italy

In their seminal and well­known article, Pintrich et al. (1993; Pintrich, 1999) suggested the need to go beyond “cold” conceptual change and to consider the affective, motivational, and situational factors that may affect knowledge revision. Among the motivational factors, Pin­trich and colleagues included epistemic beliefs, together with other, more typical motivational constructs, such as mastery goals, interest, importance, utility value, self­efficacy, and control beliefs. In describ­ing his vision of conceptual change through five propositions, Pin­trich (1999, p. 37) stated in the second that the “Adoption of more ‘constructivist’ epistemological beliefs should facilitate conceptual change.” Pintrich asserted, in particular, that belief in knowledge as complex and continuously evolving acts as a resource in the process of knowledge revision. In contrast, belief in knowledge as simple and certain places constraints on the potential for conceptual change as students tend to foreclose their thinking and do not consider alterna­tive views. Since the mid 1990s research on the relationship between epistemic beliefs and conceptual change has begun to empirically sup­port Pintrich’s proposition, although most (if not almost all) stud­ies have been carried out with college or high school students (e.g., Mason and Boscolo, 2004; Sinatra et al., 2003; Qian and Alverman, 1995, 2000; Stathopoulou and Vosniadou, 2007b; Windschitl and Andre, 1998).

In this chapter the relationship between epistemic beliefs and inten­tional change of knowledge is focused on elementary and middle school students. Addressing learners in grades one to eight appears to be rel­evant for two main reasons. First, their beliefs about academic knowl­edge (Buehl and Alexander, 2001) – that is, knowledge regarding the disciplines they encountered at school – should be relatively recently formed, since they have several years’ experience in the educational

Lucia Mason 259

context, so it is possible to document the influence of the first con­structed representations about knowledge and knowing. Second, it can be documented how critical the role of epistemic beliefs is, given their early impact on the construction and re­construction of new knowledge by students who are required to learn disciplinary content through lec­tures and text study.

Before proceeding, a clarification of terminology is required. In the literature, both the terms epistemological and epistemic are used to refer to beliefs about the nature of knowledge and the process of knowing, although the former to a greater extent. Recently, for the purpose of clarity, some scholars (Murphy et al., 2007; Alexander and Sinatra, 2007) proposed distinguishing beliefs about knowledge and knowing (epistemic) from beliefs about the study of knowledge (epistemological). In fact, Kitchener (2002) had already argued in favor of this distinc­tion. In accordance with these scholars, it should be acknowledged that research in educational, as well as in developmental, psychology deals with epistemic beliefs since it concerns personal, implicit or explicit, assumptions about knowledge, and not about the study of knowledge. It is most likely that students have epistemic beliefs. Consistently, in this chapter the adjective epistemic is used since it is focused on stu­dents’ personal beliefs about the nature, source, and justification of knowledge. However, it should also be acknowledged that psychology researchers who examine individuals’ epistemic beliefs are engaged in research about epistemological beliefs (Vosniadou, 2007).

As pointed out by Bendixen (2002), who examined epistemic doubt, it is also important to note that epistemic beliefs are not “cold” cogni­tion. The affective nature of deeply held beliefs is an important aspect to be taken into consideration in teaching and learning processes. The recognition of uncertainty and complexity of knowledge can be worry­ing for students, as well as having affective consequences (i.e., feelings of fear, confusion, or anxiety) which may derive from abandoning the security of absolutistic stances (Hofer, 2005).

The purpose of this chapter is twofold: to argue that epistemic beliefs can act as either resources or constraints in the process of conceptual change, and to examine why they affect it. In doing so, I review empiri­cal research that involves younger students (grades one to eight). In addition, I examine students’ learning regarding science domains, the subject areas in which the relationship between epistemic beliefs and conceptual change has been investigated.

The chapter is divided into six sections in order to address this pur­pose. In the first section, I describe recent models of the process of conceptual change, which deal with the subtle interaction between the

Beliefs about knowledge and revision of knowledge260

multiple factors that shape it. Within these models epistemic beliefs can play a role. In the second section, I focus on the main issues of research on younger students’ epistemic beliefs arising from developmental psy­chology, educational psychology, and science education. In the third sec­tion, I review empirical studies about the influence of epistemic beliefs on knowledge change in grades one to eight. By referring to the notion of intentionality in learning, the role of epistemic beliefs is introduced in the fourth section. On the basis of the limited research available, it is argued that it may affect knowledge restructuring both directly and indirectly. Examples of data from our empirical research with elemen­tary and middle school students illustrate and clarify the arguments being proposed. Finally, the fifth and sixth sections close the chapter with educational implications and directions for future research.

Recent models of conceptual change

Prior to Pintrich et al.’s (1993) article, most research on knowledge revision – which made a tremendous contribution to the field – focused mainly on students’ existing knowledge structures (e.g., Chi, 1992; West and Pines, 1985), developmental changes (e.g., Carey, 1985; Vos­niadou and Brewer, 1987, 1992), and teaching strategies for promot­ing knowledge revision (e.g., Posner et al., 1982). As noted, Pintrich et al.’s article inspired a “warming trend” of research that attempted to change, or at least to widen, the focus of investigation into conceptual restructuring (Sinatra, 2005). Current research highlights the intri­cacy of the change process and the interplay of multiple factors that shape that process (Sinatra and Mason, 2008; Murphy and Mason, 2006).

In educational psychology two models have been proposed as attempts to explain the multiple interrelations among cognitive, affective, and contextual factors that underlie the process of knowledge revision. Both are dual­process models (Stanovich, 1999) since the cognitive architec­ture assumed in these models is automatic, or heuristic (i.e., low cogni­tive engagement), and intentional (i.e., high cognitive engagement).

Cognitive reconstruction of knowledge model (CRKM). Dole and Sinatra (1998) were the first scholars to propose an integrated model of con­ceptual change processes, which was intended to go beyond a merely cognitive or rational perspective and to take into account a multiplicity of factors that underlie knowledge revision. Specifically, the reference to social psychology led the scholars to examine affective factors under­lying the change in representations. According to CRKM, the inter­play of learner and message characteristics generates a certain degree

Lucia Mason 261

of engagement, that is, the likelihood of change. To facilitate this, the message (i.e., a new conception) needs to be comprehensible, plausible, coherent, and rhetorically compelling. These features interact dynami­cally with the individual characteristics of a student. Learner variables include background knowledge and motivational factors. An existing conception can be analyzed in terms of strength, coherence, and com­mitment. All three aspects should be considered from a motivational point of view. If existing ideas are conceptually strong and stable, and a high degree of commitment is attached to them, for example, in relation to an epistemic belief in knowledge as absolute and certain, the learner resistance to conceptual change is high. Besides background knowl­edge, motivational factors also contribute to determine the likelihood of change. Several motivational components have been included in the model, such as dissatisfaction with one’s ideas, personal relevance (e.g., deriving from interest in the topic, investment in the outcome, self­ efficacy, etc.), and personal traits that sustain engagement, like the need to be involved in effortful thinking. In addition, CRKM includes the social context as a factor that potentially supports or undermines students’ motivation.

The change described in this model is not a linear process but rather a dynamic and iterative one. The interaction between the learner and message characteristics generates a degree of engagement. Enduring conceptual change is likely to occur only if learners are highly engaged at a metacognitive level, so that they process deeply the new content using elaborative strategies and making substantive reflections. Learn­ers can produce a temporary change in conceptions if, for example, the social context of a group or classroom pressures them to see a phenom­enon in a different way. It may also be the case that reading a particu­larly attractive text leads learners to generate a weak change. However, if students’ cognitive engagement is rather low, they will more likely return to the original representations, which have never been com­pletely abandoned (Dole and Sinatra, 1998).

Cognitive-affective model of conceptual change (CAMCC). To follow Pintrich et al.’s (1993) recommendation, Gregoire (2003) has proposed a dual­process of conceptual change, which was also inspired by the literature on belief and attitude change. The model was intended to pro­vide an account of why teachers’ subject­matter beliefs are resistant to change, so that they do not adopt reform­oriented curricula which con­trasts with them. According to the author, this model has been devised to be “truly a hot model of conceptual change” (original emphasis, p. 163) as more motivational factors than the CRKM have been taken into account. These also include teachers’ fears and confidence levels, which

Beliefs about knowledge and revision of knowledge262

affect receptiveness to a given message. In addition, CAMMC takes into account individuals’ goals, as well as their prior beliefs. If teachers per­ceive a message regarding educational reform as threatening, it is more likely that they adopt an avoidance rather than an approach goal. In con­trast, if they perceive the message as challenging, they may formulate an approach goal. Furthermore, Gregoire’s (2003) model describes the appraisal process that takes place automatically prior to the processing of the message, which generates affective responses.

As in Dole and Sinatra’s (1998) model, enduring change cannot occur when teachers are not involved in systematic processing, although the latter does not guarantee the former.

Although Gregoire’s (2003, p. 116) model claims to be “truly a hot model,” thus having a higher temperature than Dole and Sinatra’s model, both models focus on the interaction between learners’ cognitive and motivational factors (prior knowledge, processing ability, self­efficacy) and the characteristics of a message in a given context, which underlie the change in representations (Sinatra, 2005). In addition, both models give space to the consideration of epistemic beliefs as factors that may have a critical role in knowledge restructuring. Sinatra et al. (2003) and Gregoire Gill et al. (2004), in fact, articulated their theoretical speculations and carried out empirical studies providing evidence of the influence of beliefs about knowledge and knowing on conceptual change in college students. In a study about their understanding and acceptance of three scientific topics varying in degree of controversy – human evolution, animal evolution, and photosynthesis – it emerged that epistemic beliefs and cognitive dispositions predicted acceptance of scientific knowledge only for the first topic, the most controver­sial. Students who were more likely to accept human evolution were those who believed more in knowledge as complex and uncertain, and who enjoyed more critical and open­minded thinking (Sinatra et al., 2003). The relationship between general epistemic beliefs and the change process was also documented in a study with preservice teach­ers, in which both their general and subject­specific (mathematics) beliefs about knowledge and knowing were assessed. Findings revealed that the former were related to the change of participants’ views about teaching and learning in mathematics (Gregoire Gill et al., 2004).

Considered as important factors that may influence knowledge revi­sion by most recent models of conceptual change, epistemic beliefs have received increasing attention in research on learning processes. In the next section the main outcomes of investigations about the nature and role of this type of beliefs in younger students are introduced, which have been carried out within different research traditions.

Lucia Mason 263

Epistemic beliefs in younger students: research traditions and main findings

Beliefs about the nature and source of knowledge, its truth value and justification criteria of assertions, namely epistemic beliefs, are the focus of a research area that has its origins in Perry’s (1970) pioneering work. Studies on the epistemic beliefs of children and young adoles­cents, as well as older individuals, have been conducted within three main fields: (1) developmental psychology; (2) educational psychology; and (3) science education. It is not within the aim of this chapter to review all the available literature, in which the same psychological phe­nomena are often described using different terminology (Hofer and Pintrich, 1997), but rather to introduce the main issues in these fields that are relevant to the research on the relationship between epistemic beliefs and conceptual change.

(1) In the developmental psychology literature, epistemic thinking is conceived in terms of cognitive structures comprising coherent and integrated representations, which may feature a level or stage of cogni­tive development. A shared developmental progression can be identified across the authors’ models, which are mainly based on interview stud­ies. In Kuhn’s (2000; Kuhn and Weinstock, 2002) terms, individuals shift from a realist to an absolutist to a multiplist to an evaluativist point of view. From a realist position a child (typically at age three) believes that people’s assertions mirror objective reality, which cannot be inac­curately rendered. The shift from realism to absolutism implies moving from unreflected knowing about reality to metacognitive reflection on one’s and others’ knowledge assertions. From the absolutist view, knowl­edge is absolute, certain, non­problematic, right or wrong, and does not need to be justified since it is based on observations in reality or author­ity (prevalence of the objective dimension). This stance is most com­mon in children but can also be found in adolescents and adults. From the multiplist position, knowledge is conceived as ambiguous and idi­osyncratic, so that each individual has his or her own views and truths. Knowledge, in fact, becomes an opinion (prevalence of the subjective dimension). The emergence of skeptical doubt is typical of adoles­cents (Chandler, 1987, 1988). At the evaluativist level, which develops until adulthood, an individual believes that there are shared norms of inquiry and knowing, and some positions may reasonably be more sup­ported and sustainable than others. Only at this level are the objective and subjective dimensions balanced, as they are integrated and coordi­nated, and one does not dominate the other (Kuhn, 2000; Kuhn and Weinstock, 2002). It is worth noting that within an integrated model of

Beliefs about knowledge and revision of knowledge264

cognitive development, Kuhn (2000) situates epistemic understanding at the third level of meta­knowing (“How does one come to know?”; “What do I know of what I know?”), following the cognitive (“What do I know?”) and strategic (“How do I know it?”) levels.

In a study by Kuhn et al. (2000), in which epistemic understand­ing has been measured by means of a paper­and­pencil instrument devised to identify absolutist, multiplist, and evaluativist positions, fifth­ graders differed from undergraduates regarding the first (from absolutism to multiplism) and the second (from multiplism to evalua­tivism) transitions. The fifth­graders abandoned absolutist positions less frequently than undergraduates in the first transition, as well as multiplist positions in the second. A study with Italian fifth­, eighth­, eleventh­, and thirteenth­graders, using the same instrument, indi­cated that a critical step in the development of epistemic thinking occurs with the shift from elementary to middle school when it is more likely that conflicting claims are accepted as legitimate inter­pretations of phenomena and events (Mason, Boldrin, and Zurlo, 2006).

(2) Research in educational psychology has flourished after a very influential article by Schommer (1990), in which epistemic beliefs were conceived as a set of more or less independent beliefs about the nature and acquisition of knowledge, organized around dimensions. Following this perspective, researchers have conducted investigations by means of questionnaires that ask for the level of agreement with particular items, measured on a Likert­type scale, rather than through interviews. Schommer’s (1990) Epistemological Questionnaire, made up of sixty­three items, became the most commonly used instrument in research aimed at examining the relationship between domain­general beliefs about knowledge and learning outcomes. Modified and/or shorter ver­sions of this questionnaire, based on criticism of the original, were developed (e.g., Schraw et al., 1995).

It should be acknowledged that within educational psychology, through research with college students, epistemic beliefs have also been conceptualized as theories with a core of multiple dimensions of beliefs (Hofer, 2000). This means that individuals’ views are not to be seen as a collection of independent ideas, but rather as a coherent integration of compatible perspectives at both the domain­general and domain­specific levels. It is supposed that an individual may hold a theory about knowledge in general as well as theories about knowledge in domains such as math or history (Buehl et al., 2002; Muis et al., 2006). Both general and domain­specific theories operate as organized views rather than as a sum of single beliefs.

Lucia Mason 265

Despite divergences in conceptualizations, there is agreement on the dimensions that comprise epistemic beliefs (Hofer, 2000, 2004). The two main belief areas, according to the philosophical definition of epis­temology, are beliefs about the nature of knowledge (what knowledge is) and beliefs about the process of knowing (how one comes to know). There are two dimensions within the first area (knowledge):

• certainty of knowledge: the degree to which knowledge is conceived as stable or changing, ranging from absolute to tentative and evolving knowledge, andsimplicity of knowledge• : the degree to which knowledge is conceived as compartmentalized or interrelated, ranging from knowledge as made up of discrete and simple facts to knowledge as complex and inter­related concepts.

There are also two dimensions identifiable within the second area (knowing):

• source of knowledge: the relationship between knower and known, ranging from the belief that knowledge resides outside the self and is transmitted, to the belief that it is constructed by the self, andthe • justification of knowledge: what makes a sufficient knowledge claim, ranging from the belief in observation or authority as sources, to the belief in the use of rules of inquiry and evaluation of expertise.

Scholars interested in the role of epistemic beliefs on conceptual change in older students, which were measured by questionnaires, examined the impact of the various dimensions emerging from factor analyses.

Are epistemic beliefs in younger students so differentiated along mul­tiple dimensions to be identified by means of a questionnaire? Schom­mer­Aikins (2002) posited that they are undifferentiated in early life. However, through culture, parents, peers, and education epistemic beliefs gradually differentiate among the different aspects of knowl­edge. As described in more detail below, there is evidence that fifth­graders’ domain­specific beliefs have been measured by a self­report questionnaire aimed at identifying beliefs about four dimensions of scientific knowledge (Conley et al., 2004; Mason et al., in press). In addition, shorter versions of Schommer’s (1990) questionnaire have been used with over 1,200 students in middle school (Mason and Gava, 2007; Schommer et al., 2000) revealing a three, instead of four, fac­tor­structure of epistemic beliefs. More specifically, the conceptually distinct dimensions of simplicity of knowledge and certainty of knowl­edge merged into a single dimension, “simple and certain knowledge.”

Beliefs about knowledge and revision of knowledge266

Younger students may have a more limited and simplistic view of knowl­edge per se (Schommer, 2002). Nevertheless, they have an epistemic understanding that impacts learning processes, including conceptual change, as it is argued below.

(3) If the studies in the two previously mentioned research traditions dealt mainly with domain­general beliefs about knowledge and know­ing, studies in science education examined domain­specific epistemic beliefs. To describe the main features of the representations about sci­entific knowledge and knowing arising from the few studies that have involved participants in grades one to eight through interviews, we refer to Carey et al.’s (1989) outcomes regarding seventh­graders’ beliefs about scientists’ ideas, experiments, and data as an example. The authors iden­tified three general levels of responses:

Level 1: Students do not show a clear distinction between scientists’ •ideas and activities. They believe that an activity is performed to achieve the activity itself and that the goal of science is to discover facts about reality and invent new things.Level 2: Students clearly differentiate ideas from activities. They •believe that an experiment is performed to verify the rightness of an idea and that refusing or revising an idea is possible on the basis of the data. However, students do not recognize that a revised idea cov­ers old and new data and believe that the goal of science is to under­stand how things work.Level 3: As in the previous level, students clearly differentiate between •ideas and experiments and believe that an experiment is aimed at verifying or exploring ideas, and that unexpected results influence the development of ideas. They also understand the cyclic nature of science and believe that the goal is to construct finer explanations of events and phenomena.

More specifically, for instance, a hypothesis is conceived as an idea or a guess at the first level of epistemic understanding. At the second level it is still perceived in terms of an idea or guess, but also as some­thing that can be tested, which is clearly related to an experiment or phenomenon. At the third level, a hypothesis is also believed to be a tool for interpreting the results of an experiment, and is developed on the basis of those results. It can be said that the developmental trend is rep­resented by the progress from a naïve objectivist toward a sophisticated constructivist view of science.

Fifth­graders’ scientific epistemic beliefs have also been measured by means of a questionnaire aimed at examining what they think about the purpose, changeability, and coherence of science, the role of experiments

Lucia Mason 267

in developing scientific knowledge, and the source of science knowledge (Elder, 2002). It emerged that they held some sophisticated representa­tions and some naïve beliefs. To give an account of this outcome, the author hypothesized that fifth­graders may be receptive to educational interventions implemented to foster specific aspects of their views of science. In a more recent study, fifth­graders’ beliefs about science and scientific knowledge were measured by means of a questionnaire, sub­stantially based on Elder’s (2002) self­report instrument. Evidence of the hypothesized four­dimensional structure of epistemic beliefs – source, certainty, development, and justification of scientific knowledge – was provided by a confirmatory factor analysis (Conley et al., 2004).

In studying the relationship between epistemic beliefs about science and learning science, Songer and Linn (1991; Linn and Songer, 1993) documented that eighth­graders with a dynamic view of science (e.g., developing and changing) learned more integrated knowledge than those with a static view (e.g., certain and stable), showing an understanding of the principles underlining many thermodynamic phenomena.

Influence of epistemic beliefs on conceptual change

Do beliefs about knowledge influence the revision of knowledge in younger students? Although few, some studies in the educational psychology literature document the influence of elementary and middle school students’ epistemic beliefs on conceptual change processes.

Mason (2000) investigated the role of epistemic beliefs in relation to anomalous data in theory change on two controversial topics. In this regard, Chinn and Brewer (1993, 1998) pointed out that anomalous data are a key component of many instructional methods of a construc­tivist nature for fostering knowledge change in science. However, the authors also stated that anomalous data often fail to promote concep­tual change as students may discredit or discount them. The taxonomy of various possible responses to anomalous data as mediating belief change provided by Chinn and Brewer (1993, 1998) includes ignoring, rejecting, excluding them as an explanation for a given phenomenon, keeping them in abeyance, and reinterpreting them to make peripheral or central change in currently held theories. Beliefs about the nature of knowledge in general, or related to a specific field can, to some extent, facilitate or impede anomalous data acceptance, which, in turn, can lead to knowledge revision. Therefore, understanding if students’ epis­temic beliefs influence their responses to conflicting information, and why they respond as they do, appears to be a crucial issue to under­standing theory change (Mason, 2000).

Beliefs about knowledge and revision of knowledge268

Two controversial topics were introduced to 126 students in eighth­grade. One was scientific, the extinction of dinosaurs in the Cretaceous era, and the other historical, the construction of the great pyramids in Giza (Egypt). Two theories for each topic were presented, the first was familiar and the second alternative. Introduction of the latter was pre­ceded by the presentation of supporting evidence, but which conflicted with the familiar theory. Findings indicated that acceptance of anomal­ous data made the most significant contribution to theory change. The more the anomalous data were considered to be valid and incoherent with the theory held by the eighth­graders, the more they accepted the alternative theories. Students who discounted conflicting data, either by evaluating them as invalid or consistent with their familiar theory, were more likely to refuse the alternative theories. Interestingly, for the scientific topic acceptance of anomalous data was related, although modestly, to students’ belief about the stability (certain/evolving) and source (handed down by authority/derived from reason) of knowledge. Both these aspects of the nature of knowledge made up one of the four dimensions that emerged from an exploratory factor analysis of the data collected using Schommer’s (1990) questionnaire. This was the only dimension of students’ belief network that was taken into consideration as the most pertinent to the focus of the study. Students who tended to accept evidence conflicting with their conceptions and, consequently, to change their theory about dinosaur extinction were those who believed more in the changing nature of knowledge derived from reason. Their epistemic belief mediated the acceptance of evidence conflicting with their prior knowledge, although this effect was not strong.

A qualitative analysis of the reasons given by the students to justify their initial theory preference, as well as their acceptance or rejection of the anomalous data discounting it, revealed that they widely appealed to their epistemic belief in the authoritative source of knowledge. To illustrate the epistemic nature of many justifications given, four exam­ples are provided below. The first two are based on the authoritative source of school, teacher, school books, and scientists’ work. They were functional to the students evaluating the familiar theory about dinosaur extinction as completely true (Mason 2001, pp. 463 and 465):

I believe this theory is true as it says what I have studied at school; the same things. Scientists have done many investigations which satisfy me. I have never heard of other theories on that. (Eloisa)

I have always trusted books. I have studied this theory at school, my history teacher taught it. (Fabio)

It is interesting that doubts in the authoritative source of knowledge were also referred to for not accepting an alternative theory, which was

Lucia Mason 269

unfamiliar to the students, or for not accepting the anomalous data pro­vided to discount the known theory. The following example expresses an epistemic reason underlying theory rejection, which is based on a methodological consideration regarding knowledge justification (Mason, 2001, p. 466):

I believe this theory is not true because the calculations that computers make are not always right. (Alessia)

The reliability of scientists’ work was also questioned by students to doubt or reject the validity of data conflicting with their prior knowl­edge, as illustrated by the following justification, which was focused on the conjectural nature of scientific knowledge and the fallibility of scientists’ activity. In this case, this sophisticated epistemic view was functional to the student not changing her initial theory on dinosaur extinction (Mason, 2001, p. 467):

I am not sure that scientists speak the truth. They are human beings and so they can make mistakes. They have only hypotheses and are forced to make suppositions about timing the layer of clay. Yes, they might have used spe­cific techniques but they do not know the real cause of the phenomenon. (Pamela)

All these reasons given by students to motivate their positions about the familiar or unfamiliar theory and the data conflicting with the former, illustrate the activation of epistemic beliefs in the context of theory evaluation. In particular, two epistemic dimensions of the know­ing process, that is, beliefs about the source of knowledge and justifica­tion of knowledge, clearly emerge.

More data on the role of epistemic beliefs in relation to anomalous data derive, albeit indirectly, from recent systematic research carried out by Chinn and Malhotra (2004). The authors expected that conceptual change in response to the introduction of data conflicting with prior knowledge could be potentially impeded at any of four processes involved: observa­tion, interpretation, generalization, or retention. Four hundred students in fourth, fifth, and sixth grades experienced difficulties observing accu­rately but, at the same time, did not make observations according to their expectations. Since they took into account their observations as the basis for their conceptions, the authors evinced the presence of an implicit epis­temology in them, which made them distinguish between theory and evi­dence, and change conceptions in response to their observations. When upper elementary school children made correct observations, they were able to change their initial conceptions to respond to the empirical data in an epistemically consistent way, but if they did not make the correct observations, they were impeded in changing them.

Beliefs about knowledge and revision of knowledge270

The effect of epistemic beliefs on changing conceptions about bio­logical evolution in 110 eighth­graders in relation to an instructional variable, the type of text to be read to learn new knowledge, has been the focus of a study by Mason and Gava (2007). Beliefs about know­ledge and knowing were measured using a shorter version of Schom­mer’s questionnaire (thirty­six items). Three dimensions emerged from the factor analysis (as mentioned above), but only scores for one dimension (beliefs in certain and simple knowledge) were used for sub­sequent analyses, as this dimension was most relevant to the study. Two experimental conditions were examined. In the experimental condition students read a refutational text, while in the control con­dition they read a normal textbook expository text. A refutational text is one which directly states students’ alternative conceptions about a topic, refutes them, and presents the scientific conceptions as viable alternatives (e.g., Alverman and Hynd, 1989; Hynd, 2003). Therefore, not only do refutational texts activate the readers’ alternative prior knowledge but they also make explicit the inconsistencies between preexisting knowledge and the new knowledge. The same concepts of biological evolution were introduced in both texts, but the former activated and challenged students’ alternative conceptions about the scientific phenomena to be learned. Within each condition there were students with more (i.e., knowledge as uncertain and complex) and less sophisticated (i.e., knowledge as certain and simple) epistemic beliefs.1 Learners’ conceptions were ascertained at three testing times: pre­test, immediate posttest, and delayed posttest. Findings revealed that both the learner factor, epistemic beliefs, and the instructional fac­tor, the structure of the text to be read to learn new concepts, affect knowledge revision, as reflected in the students’ explanations. Holding more sophisticated beliefs about the nature of knowledge (Figure 9.1), as well as reading the refutational text, led to a greater change in the initial conceptions.

Interestingly, a significant interaction between the two factors also revealed that students who believed more in the changing and complex

1 In the epistemic beliefs literature, terms such as “naïve” and “sophisticated” or “less advanced” and “more advanced” are commonly used. We retain them for reasons of clarity and not to attribute any unfortunate connotation to the terms. In accordance with Schommer­Aikins (2002), who suggested balance, we think that when an indi­vidual holds sophisticated beliefs, it means that he or she believes mostly in complex and tentative knowledge. However, at the same time, the individual may believe that there is knowledge that is stable and/or isolated. What differentiates an advanced epi­stemic thinker from a less advanced one is that the belief in certain and simple know­ledge is predominant in the latter and the exception in the former.

Lucia Mason 271

nature of knowledge where those who benefited more from reading the refutational text that stated and challenged their alternative con­ceptions. Nevertheless, reading the refutational text compensated less mature epistemic beliefs. Furthermore, both independent variables affected learners’ metaconceptual awareness of changes in their own conceptions (Figure 9.2).

Students who had more advanced representations of knowledge and who read the text that explicitly challenged their alternative concep­tions showed greater awareness of the changes in their conceptual structures, as can be seen in the following three examples (Mason and Gava, 2007):

Before I believed that animals changed because they wanted to adapt to the environment. Now I think that animals cannot change their characteristics because they must fit the environment, but that through the natural selection only those animals that get some advantages for survival, survive. (P2)

Before reading the text I thought like Lamarck, that the species acquire new characteristics by the use of organs, and then they transmit them. Now I know that at the basis of a speciation there is a mutation. (P122)

Before I believed that giraffes had developed a long neck to reach the high trees and to be able to eat. Instead, there was a random genetic mutation and the giraffes with a longer neck survived thanks to this characteristic. (P25)

Lessadvanced More

advanced

012345

6

7

8

9

10

Pretest

Posttest 1

Posttest 2

Figure 9.1: Mean scores for conceptual change about biological evolu­tion by epistemic beliefs in eighth­graders (scores are adjusted by the covariate of reading comprehension)

Beliefs about knowledge and revision of knowledge272

The emergent metacognitive awareness expressed in these examples, which was stimulated by epistemic beliefs, and by reading the refuta­tional text, probably contributed to their greater degree of conceptual change.

In a study on revising conceptions about the nature and diffusion of light, the impact of ninety­four fifth­graders’ epistemic beliefs was investigated in relation to the instructional text type and also in relation to topic interest (Mason et al., 2008). An energizing role in cognitive processing is attributed to this motivational variable, as it stimulates attention arousal, positive emotional reactions, effort, and willingness to persist in the task (e.g., Ainley et al., 2002). Deep cognitive process­ing of the content to be learned is therefore sustained by high interest which is beneficial to knowledge revision, as it requires a high level of engagement.

In this study, domain­specific beliefs about knowledge and knowing were measured, that is, science­related beliefs. More specifically, two of the four scales that comprise Conley et al.’s (2004) questionnaire were taken into consideration: certainty and development of scientific knowl­edge. In this study, too, participants were divided into two experimental conditions. In one they read a normal scientific expository text about light, vision, and color; in the other they read a refutational text on the same topic. They were also tested three times, from pre to delayed posttest. In accordance with the most recent research on conceptual

00.10.20.30.40.50.60.70.80.91.0

Lessadvanced

Moreadvanced

Lessadvanced

Moreadvanced

Refutationaltext

Non-refutationaltext

Figure 9.2: Mean scores for metaconceptual awareness at immediate posttest by epistemic beliefs and type of text in eighth­graders (scores are adjusted by the covariate of reading comprehension)

Lucia Mason 273

change (Murphy and Mason, 2006; Sinatra and Mason, in press), over­all findings revealed that knowledge revision was affected by several interactions between the factors examined. Learners with high interest in the topic of light outscored students with low topic interest at both immediate and delayed posttests. In addition, students who received information from the refutational text were facilitated in conceptual change much more than those who read the traditional text. Further­more, an interaction between all three variables emerged. At both post­tests the highest scores for conceptual change, as reflected in students’ explanations, were obtained by learners with more advanced epistemic beliefs, high topic interest, and who read the refutational text, although these scores decreased from the immediate to the delayed posttest. In addition, in this study too, the text structure factor produced a com­pensatory effect for learners with low topic interest and less advanced beliefs about scientific knowledge at both immediate and delayed post­tests (Figure 9.3).

To illustrate the change in conceptions occurring in the fifth­ graders’ conceptual structures, two examples of explanations at pre­ and posttests follow. The first example shows the answers to the question “In this diagram there is a candle. Draw the light of the can­dle to show what happens when the candle burns. Explain what you have drawn.”

Pretest: I have drawn the light which is still and illuminates a specific point.

Immediate posttest: I have drawn the light rays that do not meet any obstacle so they diffuse in all directions (P96, refutational text reading, high topic inter­est, and more advanced epistemic beliefs).

The second example reports the explanations given by a participant to the question “A child is looking at an orange on the table. Explain why she sees the orange as an orange color.”

Pretest: I have drawn an arrow to show that the child is looking at the orange. She sees it as orange because its skin is orange.

Delayed posttest: The light, which is made up of the different colors of the rain­bow, absorbs all colors but the orange, so the child sees the orange as orange (P43, refutational text reading, low topic interest, and more advanced epis­temic beliefs).

In another study with seventy fifth­graders, the impact of science­related beliefs on conceptual change about magnets was examined in relation to their low or high mastery and performance goal orienta­tions (Mason, Boldrin, and Vanzetta, 2006). Domain­specific epis­temic beliefs were measured using the entire Conley et al. (2004)

Beliefs about knowledge and revision of knowledge274

questionnaire including beliefs about the source, certainty, develop­ment, and justification of knowledge. According to the literature, mas­tery goals, with their focus on understanding, imply attentional focus on the task at hand, which is associated with systematic information processing and this, in turn, produces higher outcomes (e.g., Maehr, 1984; Linnenbrick and Pintrich, 2002; Smiley and Dweck, 1994). Therefore, students who adopt a mastery goal should be facilitated when the learning task requires them to radically change their concep­tions, as this change relies on deep cognitive elaborative strategies. In this study a significant negative correlation between epistemic beliefs and performance goal emerged (r = –.31, p < .01): the more they had

Low

int a

nd le

ss a

dv b

el

Low

int a

nd m

ore

adv

bel

Hig

h in

t and

less

adv

bel

Hig

h in

t and

mor

e ad

v be

l

Low

int a

nd le

ss a

dv b

el

Low

int a

nd m

ore

adv

bel

Hig

h in

t and

less

adv

bel

Hig

h in

t and

mor

e ad

v be

l

0

2

4

6

8

10

PretestPosttest 1Posttest 2

Refutationaltext

Non-refutationaltext

Figure 9.3: Overall mean scores for conceptual change about light showing the interaction between epistemic beliefs, topic interest, and text type in fifth­graders (scores are adjusted by the covariate of read­ing comprehension). From “On warm conceptual change: The in­terplay of text, epistemological beliefs, and interest,” by L. Mason, M. Gava, and A. Boldrin (2008), Journal of Educational Psychology. Copyright 2008 by American Psychological Association. Reprinted with permission

Lucia Mason 275

sophisticated beliefs about scientific knowledge and knowing, the less they were performance oriented. In addition, an interesting interaction emerged between epistemic beliefs and mastery goals. When learners had less sophisticated science­related beliefs, they benefited more from being highly focused on understanding and learning the new scientific content. In contrast, if their science­related epistemic beliefs were more sophisticated, mastery goal orientation did not make a difference to the change of explanations about magnets (Figure 9.4).

In sum, these studies highlight the impact of epistemic beliefs in knowledge revision processes, either per se or in relation to other fac­tors that have the potential to support changes in the understanding of disciplinary concepts. Regardless of whether they are domain­general or domain­specific, more constructivist beliefs (knowledge as complex, developing, changing, and constructed by the self) act as a powerful resource for conceptual change. If learners hold more naïve epistemic beliefs (i.e., knowledge as simple, stable, certain, residing outside the self, and transmitted by authority), they may be helped to revise their alternative conceptions when other motivational resources are avail­able to them, such as topic interest or mastery goal orientation, which can compensate for the disadvantage of less advanced beliefs about knowledge and knowing.

Low

mas

tery

Hig

h m

aste

ry

Low

mas

tery

Hig

h m

aste

ry

Pretest

Posttest

0

2

4

6

8

10

Figure 9.4: Mean scores for conceptual change about magnets by epistemic beliefs and mastery goal orientation in fifth­graders (scores are adjusted by the covariate of reading comprehension)

Beliefs about knowledge and revision of knowledge276

Why do epistemic beliefs affect conceptual change?

Pintrich (1999) was the first scholar to speculate that believing in knowl­edge as certain, absolute, simple, and comprising many isolated facts to be memorized, or believing in knowledge as a complex system of inter­connected elements that are continuously evolving and must be learned by integrating information, can make a difference in conceptual change learning. Empirical research has now documented that even young stu­dents’ beliefs about knowledge, which may have developed relatively recently, influence the revision of their knowledge. In this regard, the current issue is: why do more sophisticated epistemic beliefs act as a resource while less sophisticated beliefs act as a constraint in conceptual change? To provide an account within a theoretical framework in which intentionality is a key construct, Mason (2003) argued that epistemic beliefs may have a direct effect on knowledge restructuring by influenc­ing the intentions that guide a learner and, consequently, the cognitive activities that he or she performs. According to Sinatra and Pintrich (2003, p. 6), intentional conceptual change requires “goal directed and conscious initiation and regulation of cognitive, metacognitive, and motivational process to bring about a change in knowledge.” The dual­process models of conceptual change (Dole and Sinatra, 1998; Gre­goire, 2003), described above, posit that although changes can occur by chance, or without awareness, only high levels of metacognitive and motivational engagement lead to deeper and longer­lasting change (Sinatra, 2005). Epistemic beliefs may or may not help students set up the intention of learning by understanding the new content (Mason, 2003). In the case of conceptual change learning, this intention implies starting the learning process from recognizing a problem of knowledge, such as the lack of or contradictory information in one’s conceptual structures. Only more sophisticated epistemic beliefs, that is, beliefs in complex, hypothetical, and evolving knowledge, are conducive to that recognition.

In addition, as recently documented by Stathopoulou and Vosniadou (2007b), epistemic beliefs may have not only a direct but also an indirect influence on conceptual change by mediating cognitive and metacogni­tive strategy use in pursuing the goal of knowledge restructuring. Only students with more sophisticated beliefs about knowledge are likely to invest effort in solving knowledge problems concerning the state of their current understanding by deeply processing the new content (Mason, 2002, 2003). Very few studies have provided evidence of the mediat­ing role of epistemic beliefs in intentional conceptual change in older students. Using a thinking­aloud methodology, a study with college

Lucia Mason 277

students showed that their epistemic beliefs about the speed of knowl­edge acquisition were related to the overall number of cognitive proc­esses they activated while reading a dual­positional text (Kardash et al., 2000). The above­mentioned study with trainee teachers documented that those who held beliefs in knowledge as simple and certain were less likely to engage in deep thinking about the content presented in a refu­tational text, and, in turn, were less likely to change their views about mathematics teaching and learning (Gregoire Gill et al., 2004). A study with Greek tenth­graders (Stathopoulou and Vosniadou, 2007a) also showed recently that the relationship between epistemic beliefs and conceptual change is mediated by the approach to learning and the selection of study strategies. Half of the examined students, who held more constructivist epistemic beliefs about physics and were able to change their preconceptions in dynamics, were those who showed a deep approach to learning through the adoption of elaborate study strategies, as revealed by interviews, thinking­aloud, and observations while undergoing the proposed tasks.

To my knowledge, no studies about elementary or middle school stu­dents’ cognitive processing in conceptual change learning have been published. To explore this issue in order to identify the cognitive proc­esses activated while reading a text, and to examine whether epistemic beliefs affect them, as well as text understanding, I recently carried out a study (Mason, 2007). It involved thirty­seven fifth­graders, part of a wider group of students, whose beliefs about scientific knowledge and knowing were measured by Conley et al.’s (2004) questionnaire. A think­aloud methodology was used to examine the cognitive proc­esses on­line, that is during their activation. Learners were asked to think aloud while they read a refutational text about the origin of the color of objects (part of the text previously used in Mason et al., 2008). As an off­line measure, that is as a task given after text reading, they were asked to write a free recall of all they had learned from the text. Each clause of the think­aloud protocols was transcribed and catego­rized on the basis of a coding scheme adapted from previous research with older students (Kendeou and van den Broek, 2007). The qualita­tive analysis revealed the following cognitive activities in which stu­dents engaged during reading: paraphrasing (responses that capture the gist meaning of text sentences); associations (responses that show that readers refer to experience related or unrelated to the text mate­rial); correct inferences (correct explanatory inferences based on readers’ prior knowledge and text); incorrect inferences (incorrect explanatory inferences based on readers’ prior knowledge and text); conceptual change strategies (responses showing that readers are metacognitively

Beliefs about knowledge and revision of knowledge278

aware of a cognitive conflict, need for change, or change in one’s con­ceptual structures). Correlational analyses revealed that students’ epistemic beliefs were positively associated with the following on­line cognitive processes: correct inferences (r = .35), associations (r = .37), and conceptual change strategies (r = .45). Epistemic beliefs were also negatively associated with incorrect inferences (r = –.46). The more sophisticated the students’ beliefs about knowledge and knowing, the more they produced correct inferences, associations, and conceptual change strategies, as well as the less they produced incorrect infer­ences. Furthermore, epistemic beliefs were also related to the free recall (r = .47) that students were asked to write to express all that they had learned from the text: the more sophisticated their beliefs, the better their recall. In addition, a hierarchical regression analysis was carried out to see which on­line cognitive processes would predict the off­line measure of text recall while the other examined variables – prior knowledge, reading comprehension, and epistemic beliefs – were con­trolled. This statistical analysis revealed that, in the first step, reading comprehension and epistemic beliefs were predictors. In the second step, positive predictors of the recall quality were reading comprehen­sion, prior knowledge, and conceptual change strategies. Incorrect inferences negatively predicted the quality of conceptual understand­ing from the text (Table 9.1).

Table 9.1. Summary of regression analysis for conceptual change learning from text in fifth-graders

Variable B SE B β

Step1

Prior knowledge .54 .39 .16

Epistemic beliefs .05 .02 .27*

Reading comprehension .97 .20 .58**

Step 2

Prior knowledge .70 .32 .21*

Epistemic beliefs .00 .02 .01

Reading comprehension .83 .17 .49**

Conceptual change strategies .38 .15 .27*

Correct inferences .04 .12 .05

Incorrect inferences .37 .13 −.32*

* = p < .05; ** = p < .001

R2 = .54 for Step 1, p < .001; R2 = .72 for Step 2, p = .001

Lucia Mason 279

Although exploratory and preliminary, these findings highlight, on the one hand, the expected role of preexisting knowledge, reading compre­hension skills, and epistemic beliefs, and, on the other, the importance of metaconceptual reflections activated during text reading, as well as the danger of incorrect inferences. More precisely, a lower number of the latter and a higher number of the former contribute significantly to understanding a text that requires conceptual change. In addition, given the small number of participants, a Sobel test was performed to further analyze the mediation effect of conceptual change strategies in relation to epistemic beliefs. The Sobel test revealed that conceptual change strategies significantly carried the influence of epistemic beliefs on text understanding (Z = 2.25, p < .05). This outcome provided some preliminary evidence that epistemic beliefs may also have an indirect influence on conceptual change, even in young students, by mediat­ing the adoption of more powerful cognitive operations for knowledge revision. The following are some examples of different metaconceptual reflections, taken from the thinking­aloud protocols of students with more advanced epistemic beliefs about science, which were positively associated with knowledge change.

The first illustrates learners’ feelings of difficulty when trying to understand a new conception:

I’m not able to figure this out. (P6)

The second reflection is an example of students’ awareness of their own ideas:

Uh … I also think like those scientists, that light is yellow or white. (P7)

The next reflection exemplifies learners’ verbalizations about the contrast between text information and their current conception:

It’s strange … I discover a new thing because light is seven colors, while I thought it was just one or two colors. (P17)

The following example is a student reflection focused on the new understanding that has been attained:

Yes, yes, now I understand why we see the color red. (P9)

The last metaconceptual expression is an illustration of students’ verbalizations about the change taking place in their own conceptual structures:

I believed that the color red was the color that belongs to an object, for example an apple. I now understand that the color we see with our eyes is the color of the light that the apple is not able to absorb. (P1)

Beliefs about knowledge and revision of knowledge280

To conclude, the strategies used to process content to be learned seem to intervene in the relationship between epistemic beliefs and con­ceptual change. A deep approach to studying a scientific text, which results in better comprehension, is an indirect or mediating effect of holding more sophisticated epistemic beliefs. At the intentional level of cognition, engagement for conceptual change combines cognitive, metacognitive, and motivational aspects of self­regulated learning. Its highest form implies not only comparing and connecting existing and new information, but also reflecting about what an individual is think­ing and why. Epistemic beliefs contribute to engagement, which, in turn, contributes to learning by conceptual change.

Educational implications

In this chapter, I argued that epistemic beliefs are important to know­ledge revision in young students. They may have a direct effect by influencing the intentions that guide learners, as well as an indirect effect by mediating the cognitive operations that characterize the pro­cess of conceptual restructuring. In this section, I synthesize theoretical and empirical issues of the literature reviewed about the relationship between epistemic beliefs and knowledge revision, and from them draw implications from an educational standpoint.

(1) Current models of conceptual change highlight the intricacy of the change process and the delicate interactions of multiple factors that shape that process. Beliefs about knowledge and knowing interact with other individual, instructional, and contextual factors with the role of either resource or constraint. Therefore, the first educational implica­tion is that students’ beliefs about knowledge and knowing should be made explicit; students should also be aware of their epistemic pre­suppositions about knowledge and not only of their knowledge about phenomena or events. Younger learners in particular are very often unaware of the epistemic beliefs which may be the basis for their alter­native conceptions. However, awareness of these beliefs is an essential condition for considering them as an object of reflection and being able to change them when they appear to be inadequate. It is not entirely clear how young students can be helped, first of all, to be aware of their epistemic beliefs. One suggestion is that they be encouraged to verbalize what they believe about knowledge and knowing when trying to learn new knowledge, a process which often implies a comparison between naïve and scientific conceptions about a phenomenon. For instance, a student may experience difficulties understanding a text which states that what scientists considered to be scientific knowledge some decades

Lucia Mason 281

ago is now unacceptable in the light of experimental evidence. The stu­dent’s difficulties may be due to an entrenched belief that knowledge is certain and stable, thus the text content does not make sense. The first step in the path that leads the student to make sense of the text is to be aware of any underlying conviction regarding the unchanging nature of knowledge over time. Teachers should help students be aware of their presuppositions about knowledge and knowing by stimulating them, for example, to reflect upon what is clear and unclear in the material to be learned, and why, giving them the opportunity to discuss implicit epistemic beliefs, which may be an obstacle to their understanding.

(2) Empirical research on conceptual change documents that only more sophisticated beliefs facilitate knowledge restructuring. Thus, the second implication that can be drawn regards the importance of fostering students’ epistemic thinking through educational practice that takes place in the natural and complex learning environment of the classroom. Revision of knowledge for meaningful learning can be promoted by refining their views about knowledge itself. If the first step is to be aware of one’s own epistemic beliefs, the next step is to start refining those beliefs that may prevent conceptual change. In sci­ence domains in particular the strong focus on data and their use as evidence in constructing valid and supported arguments can “natu­rally” provide even young students with the opportunity to activate and refine epistemic thinking. In other words, students should be engaged on the epistemic plane, which has been defined as the highest level of metaknowing (Kuhn, 2000). Scholars who have proposed models for the development of epistemic thinking (e.g., King and Kitchener, 1994; Kuhn and Weinstock, 2002) have pointed out the need for students to deal with ill­defined problems, that is, problems for which there is no single right answer, in order to practice gathering and analyzing evidence from multiple sources, and making and defending claims. Dealing with controversial issues may be an effective way to develop epistemic thinking in the classroom since it requires collecting, ana­lyzing, and comparing multiple positions to rationally evaluate or pro­duce knowledge claims (Mason and Scirica, 2006). Nevertheless, in common educational contexts there is little or no room for debate and uncertainty as students are asked to comment on and study topics intro­duced by textbooks. A textbook does not usually report on currently debated themes; instead it presents information in a non­problematic form, even when the introduction of different perspectives could make some issues easier to understand. The logical progression of scientific advances and the coherence of scientific knowledge are underlined in textbooks, whereas the controversial nature of most scientific progress

Beliefs about knowledge and revision of knowledge282

is usually omitted. For instance, in the science classroom, even when experiments are carried out to test hypotheses, they are to some extent “artificial” with respect to the defining features of scientific inquiry. They are performed by following simple protocols, often set up by the teacher, that include only certain procedures and results.

In contrast, as documented in the literature reviewed above, the use of anomalous data or refutational texts introduce, more or less directly, different perspectives about an issue. Critical examination and evalua­tion of multiple positions on a topic may help students understand and hold an evaluativist, sophisticated view of knowledge claims. They need to be supported in coming to believe that there are shared norms of inquiry and knowing, and some positions are reasonably more justified and sustainable than others. Evaluativist thinking involves active proc­esses of reflection about the aim and nature of knowledge, its source, and justification criteria, in order to be able to progressively integrate and coordinate the objective and subjective dimensions of knowing. As research regarding the SCOPE (Science Controversies Online: Part­nerships and Education) project has recently shown, teaching through debate and argumentation allows students to refine their understanding of the nature and role of debate in science, in a context of dialogue with peers and teachers (Bell and Linn, 2002). The value of controversy in the science classroom is therefore dual: on the one hand it involves epistemic thinking to deal with the nature, source, and credibility of knowledge. On the other hand, dealing with these aspects of knowledge and knowing processes means stimulating and sustaining the refine­ment of epistemic thinking. In this regard, there is evidence that more constructivist epistemic beliefs about science, for example, can be pro­moted by instruction in elementary and middle school students (Bell and Linn, 2002; Conley et al., 2004; Smith et al., 2000). However, it should be noted that controversial topics do not pertain only to science, to be valued only in the science classes. In other disciplines, such as history, there are many opportunities for students to deal with multi­ple written documents to help them learn not only new content about historical events and phenomena, but also how historical accounts are constructed, leading to an understanding of the historian’s job as a scientist. The documents present different interpretations of the same historical event while providing the opportunity to become aware that “pure” facts do not exist, and to reflect upon biased interpretations of the facts (Mason, 2002).

(3) The third educational implication highlights that fostering stu­dents’ epistemic beliefs means promoting reflective judgment or critical thinking, in King and Kitchener’s (1994) and Kuhn’s (1999) terms,

Lucia Mason 283

which is a major goal of education even for elementary and middle school students. To deal with the popular messages of science, for example, and debatable issues they will encounter in their everyday lives, students need to be able to recognize whether arguments are sup­ported by evidence, the quality and quantity of that evidence, as well as the pragmatic meaning of messages (reports or texts), that is, the intentions (Norris and Phillips, 1994). It is even more important these days for students to be able to critically evaluate information and infor­mation sources since a large amount of information is available with the simple click of a computer mouse. Internet reading implies new and different reading strategies according to the purposes of web naviga­tion. If the purpose is to search information to build new knowledge, learners should be able to use search engines not only to locate infor­mation, but also to evaluate the credibility of websites and accuracy of the information provided (Mason and Boldrin, in press), which means being reflectively and critically active on the epistemic level. There­fore, fostering students’ epistemic beliefs in the classroom is crucial to equip them, from their early years of education, with the best tools to be sophisticated consumers of information.

(4) The fourth implication regards the role of affect, which is often neglected in research on learning processes (Pintrich et al., 1993). Beliefs about knowledge and knowing have a strong affective component, as has been clearly documented (Bendixen, 2002). This means that changes can be even more challenging and demanding as students may perceive that they will experience negative emotions if they abandon their beliefs. At an early stage of epistemic development students may be strongly committed to a view of knowledge as right or wrong, certain, and trans­mitted by an omniscient authority, and feel anxiety and disorientation if this safer absolutist position has to be abandoned. The experience of epistemic doubt, which may be a mechanism of change at any stage and not only in relation to absolutism (Bendixen and Rule, 2004), can be painful and demotivating, especially for students who tend to be dependent on authority. An uncomfortable feeling of bewilderment in a sea of uncertainty may be unbearable. Affect may therefore con­strain epistemic development. However, it can also facilitate it to the extent to which doubt is accompanied by positive emotions associated with willingness to take action and responsibility, that is, volition for the resolution of doubt (Bendixen and Rule, 2004). The emotional side of epistemic development is to be taken into account to understand better how one student may engage in reflecting to move forward and come to terms with epistemic doubt, while another student may ignore feelings of epistemic doubt, or quickly revert to previous beliefs.

Beliefs about knowledge and revision of knowledge284

(5) Last but not least, the implications that concern teacher train­ing. Teachers also need to be aware of their own epistemic beliefs to be able to reflect upon them, analyze the reasons behind them, and, eventually, refine them (Woolfolk Hoy et al., 2006). Their more or less implicit assumptions about the nature of knowledge and knowledge construction in their own discipline strongly influence the way they introduce it in the classroom. Although they may be quite unaware of it, they convey to students an epistemic image of their discipline and create the epistemic climate (Bendixen and Rule, 2004) of the class­room in which teaching and learning processes take place. When they introduce the concepts of their domain, they do not only introduce knowledge but also a view about knowledge. Implicitly, day by day, they can convey – to give a simplified example – the idea that knowledge is certain, unambiguous, and possessed by recognized authorities, or that knowledge is tentative, changing in the light of new research, and validated according to shared standards of inquiry. Emphasizing one or another aspect of knowledge and knowing makes a difference. Thus, students’ epistemic belief change may start with the teacher’s epistemic belief change. A considerable part of preservice and inservice teacher training should be devoted to the “treatment” of epistemic beliefs underlying their disciplinary knowledge. Starting from an awareness of their own convictions as teachers and the implications, moving through the recognition that epistemic beliefs matter and play a subtle but sig­nificant role in teaching and learning processes, teachers can gradually produce epistemic change. I firmly believe that teachers’ beliefs (epis­temic, motivational, etc.) are a crucial factor in the educational context. Better education is still, essentially, a matter of teacher characteristics, among them their beliefs about the academic knowledge they are asked to transmit.

Future research

As a direction for future research, first there is the need to measure children’s epistemic beliefs in more appropriate and integrated ways. Interviews and self­report scales have been the instruments most widely used to identify views about knowledge and knowing. Interviews are expensive in terms of time and often cannot be used in classroom studies. Self­report scales, which can be more easily used, often have a modest reliability. Recently, other instruments, such as vignettes or scenarios, have been adopted with children (Bendixen and Haerle, 2005; Mansfield, Blythe, and Clinchy, 2002). In this regard, it should be pointed out that the main criticisms raised about most research

Lucia Mason 285

on epistemic beliefs regards the decontextualized ways in which they have been conceptualized and measured. Scholars interested in science teaching and learning (e.g., Hammer and Elby, 2002) have argued that beliefs about knowledge and knowing cannot be identified at either a general­domain or specific­domain level, but only in a given context. Their perspective is based on an alternative view of epistemic beliefs, which are not to be conceived as stable cognitive structures, consistent across contexts, that an individual does or does not have, but rather as context­sensitive cognitive resources, activated in a given situation and not in another, as revealed by classroom observations. Although I do not think that students show only epistemic inconsistencies across contexts, I agree that if we assess epistemic beliefs in a decontextualized way, we may fail to capture them appropriately, especially with young students. In accordance with Schraw (2001) and Bendixen and Rule (2004), I strongly believe that using multiple measures is preferable to using either method alone. This implies, for example, that self­report questionnaires should at least be integrated with more naturalistic tools. For this reason, we have recently begun a research program focused on examining the activation of epistemic beliefs in a particular context, that is, the context of online information searching on the world wide web (Mason and Boldrin, 2008; Mason et al., in press).

Second, efforts should be made to investigate in greater depth both the direct and indirect effects of epistemic beliefs in concep­tual change processes. The studies reviewed above document that in the literature on conceptual change there is evidence of the products which more and less sophisticated views about knowledge contribute to. Nevertheless, we do not know enough about the processes by which epistemic beliefs directly influence students’ intentions and standards to be pursued for knowledge revision, or about the factors – motiva­tional, cognitive, and metacognitive – that mediate their influence. It is important to shed more light on the explanations of the role of epis­temic beliefs in knowledge construction and reconstruction. Taking into account the specific context in which epistemic beliefs are acti­vated and measured, as well as integrating quantitative and qualitative methods of inquiry, may be steps toward a deeper understanding of how and why these beliefs can act either as resources or constraints in conceptual change.

R EF ER ENCES

Ainley, M., Hidi, S., and Berndorff, D. (2002). Interest, learning, and the psy­chological processes that mediate their relationship. Journal of Educational Psychology, 94, 545–61.

Beliefs about knowledge and revision of knowledge286

Alexander, P. A. and Sinatra, G. M. (2007). First steps: Scholars’ promising movements into a nascent field of inquiry. In S. Vosniadou et al. (Eds.), Re-framing the problem of conceptual change in learning and instruction (221–36). Oxford: Elsevier.

Alverman, D. E. and Hynd, C. (1989). Effects of prior knowledge activation modes and text structure on nonscience majors’ comprehension of phys­ics. Journal of Educational Research, 83, 97–102.

Bell, P. and Linn, M. C. (2002). Beliefs about science: How does science instruction contribute? In B. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

Bendixen, L. D. (2002). A process model of epistemic belief change. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (191–208). Mahwah, NJ: Lawrence Erlbaum Associates.

Bendixen, L. D. and Haerle, F. C. (2005). An interactive measure of children’s per-sonal epistemology. Paper presented at the 11th conference of the European Association for Research on Learning and Instruction, Nicosia, Cyprus.

Bendixen, L. D. and Rule, D. C. (2004). An integrative approach to personal epistemology: A guiding model. Educational Psychologist, 39, 69–80.

Buehl, M. M. and Alexander, P. A. (2001). Beliefs about academic knowledge. Educational Psychology Review, 13, 395–418.

Buehl, M. M., Alexander, P. A., and Murphy, P. K. (2002). Beliefs about schooled knowledge: Domain specific or domain general? Contemporary Educational Psychology, 27, 415–49.

Dole, J. A. and Sinatra, G. M. (1998). Reconceptualizing change in the cogni­tive construction of knowledge. Educational Psychologist, 33, 109–28.

Carey, S. (1985). Conceptual change in childhood. Cambridge, MA: MIT Press.Carey, S., Evans, R., Honda, M., Jay, E., and Unger, C. (1989). “An experi­

ment is when you try it and see if it works”: A study of grade 7 students’ understanding of the construction of scientific knowledge. International Journal of Science Education, 11, 514–29.

Chandler, M. J. (1987). The Othello effect: Essays on the emergence and eclipse of skeptical doubt. Human Development, 30, 137–59.

(1988). Doubt and developing theories of mind. In J. W. Astington et al. (Eds.), Developing theories of mind (387–413). New York: Cambridge Uni­versity Press.

Chi, M. T. H. (1992). Conceptual change within and across ontological cat­egories: Examples from learning and discovery in science. In R. N. Giere (Ed.), Minnesota studies in the philosophy of science, Vol. XV: Cognitive models of science (129–86). Minneapolis, MN: University of Minnesota Press.

Chinn, C. A. and Brewer, W. F. (1993). The role of anomalous data in knowl­edge acquisition: A theoretical framework and implications for science education. Review of Educational Research, 63, 1–49.

(1998). An empirical text of a taxonomy of responses to anomalous data in science. Journal of Research in Science Teaching, 35, 623–54.

Lucia Mason 287

Chinn, C. A. and Malhotra, B. A. (2004). Children’s responses to anomalous scientific data: How is conceptual change impeded? Journal of Educational Psychology, 94, 327–43.

Conley, A. M., Pintrich, P. R., Vekiri, J., and Harrison, D (2004). Changes in epistemological beliefs in elementary science students. Contemporary Edu-cational Psychology, 29, 186–204.

Elder, A. D. (2002). Characterizing fifth grade students’ epistemological beliefs in science. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (347–63). Mahwah, NJ: Lawrence Erlbaum Associates.

Gregoire, M. (2003). Is it a challenge or a threat? A dual­process model of teachers’ cognition and appraisal processes during conceptual change. Educational Psychology Review, 15, 147–79.

Gregoire Gill, M., Ashton, P., and Algina, J. (2004). Changing preservice teach­ers’ epistemological beliefs about teaching and learning in mathematics: An intervention study. Contemporary Educational Psychology, 29, 164–85.

Hammer, D. and Elby, A. (2002). On the form of personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (169–90). Mahwah, NJ: Lawrence Erlbaum Associates.

Hofer, B. K. (2000). Dimensionality and disciplinary differences in personal epistemology. Contemporary Educational Psychology, 25, 378–405.

(Ed.) (2004). Personal epistemology: Paradigmatic approaches to under­standing student’s beliefs about knowledge and knowing. Educational Psychologist, 39, 1–80 (special issue).

(2005). The legacy and the challenges: Paul Pintrich’s contributions to per­sonal epistemology research. Educational Psychologist, 40, 95–105.

Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learn­ing. Review of Educational Research, 67, 88–140.

Hynd, C. (2003). Conceptual change in response to persuasive messages. In G. M. Sinatra and P. R. Pintrich (Eds.), Intentional conceptual change (291–315). Mahwah, NJ: Lawrence Erlbaum Associates.

Kardash, C., Howell, M., and Karen, L. (2000). Effects of epistemological beliefs and topic­specific beliefs on undergraduates’ cognitive and stra­tegic processing of dual­positional text. Journal of Educational Psychology, 92(3), 524–35.

Kendeou, P. and van den Broek, P. (2007). The effects of prior knowledge and text structure on comprehension processes during reading of scientific texts. Memory and Cognition, 35, 1567–77.

King, P. A. and Kitchener, K. S. (1994). Developing reflective judgment. San Francisco, CA: Jossey­Bass.

Kitchener, R. F. (2002). Folk epistemology: An introduction. New Ideas in Psy-chology, 20, 89–105.

Kuhn, D. (1999). A developmental model of critical thinking. Educational Researcher, 28, 16–26.

Beliefs about knowledge and revision of knowledge288

(2000). Theory of mind, metacognition, and reasoning: A life­span perspec­tive. In P. Mitchell and K. J. Riggs (Eds.), Children’s reasoning and the mind (301–26). Hove, UK: Psychology Press.

Kuhn, D., Cheney, R., and Weinstock, M. (2000). The development of episte­mological understanding. Cognitive Development, 15, 309–28.

(2002). What is epistemological thinking and why does it matter? In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (121–44). Mahwah, NJ: Lawrence Erl­baum Associates.

Linn, M. C. and Songer, N. B. (1993). How do students make sense of science? Merrill-Palmer Quarterly, 39, 47–73.

Linnenbrink, E. A. and Pintrich, P. R. (2002). The role of motivational beliefs in conceptual change. In M. Limón and L. Mason (Eds.), Reconsidering conceptual change: Issues in theory and practice (115–35). Dordrecht, The Netherlands: Kluwer Academic Publishers.

Maehr, M. L. (1984). Meaning and motivation. In R. Ames and C. Ames (Eds.), Research on motivation in education: Student motivation (Vol. 1, 115–44). New York: Academic Press.

Mansfield, A. F. and Clinchy, B. M. (2002). Toward the integration of object­ivity and subjectivity: Epistemological development from 10 to 16. New Ideas in Psychology, 20, 225–62.

Mason, L. (2000). Role of anomalous data and epistemological beliefs in mid­dle students’ theory change on two controversial topics. European Journal of Psychology of Education, 15, 329–46.

(2001). Responses to anomalous data on controversial topics and theory change. Learning and Instruction, 11, 453–83.

(2002). Developing epistemological thinking to foster conceptual changes in different domains. In M. Limón and L. Mason (Eds.), Conceptual change reconsidered: Issues in theory and practice (301–36). Dordrecht, The Nether­lands: Kluwer Academic Publishers.

(2003). Personal epistemologies and intentional conceptual change. In G. M. Sinatra and P. R. Pintrich (Eds.), Intentional conceptual change (199–236). Mahwah, NJ: Lawrence Erlbaum Associates.

(2007). Effectiveness of refutational texts in relation to learner characteristics in intentional conceptual change. Paper presented at the American Educational Research Association annual meeting. Chicago, IL.

Mason, L. and Boldrin, A. (2008). Epistemic metacognition in the context of information searching on the web. In M. S. Khine (Ed.), Knowing, knowl-edge and beliefs: Epistemological studies across diverse cultures (377–404). Dor­drecht, The Netherlands: Springer.

Mason, L., Boldrin, A., and Ariasi, N. (in press). Searching the web to learn about a controversial topic: Are students epistemically active? Instructional Science.

Mason, L., Boldrin, A., and Vanzetta, A. (2006). Epistemological beliefs and achievement goals in conceptual change learning. Paper presented at the 4th EARLI SIG Symposium on “Conceptual Change,” Stockholm, Sweden.

Mason, L., Boldrin, A., and Zurlo, G. (2006). Epistemological understand­ing in different judgment domains: Relationships with gender, grade

Lucia Mason 289

level, and curriculum. International Journal of Educational Research, 45, 43–56.

Mason, L. and Boscolo, P. (2004). Role of epistemological understanding and interest in interpreting a controversy and in topic­specific belief change. Contemporary Educational Psychology, 29, 103–28.

Mason, L. and Gava, M. (2007). Effects of epistemological beliefs and learning text structure on conceptual change. In S. Vosniadou et al. (Eds.), Refram-ing the conceptual change approach in learning and instruction (165–96). Oxford: Elsevier.

Mason, L., Gava, M., and Boldrin, A. (2008). On warm conceptual change: The interplay of text, epistemological beliefs, and topic interest. Journal of Edu-cational Psychology 100, 219–309.

Mason, L. and Scirica, F. (2006). Prediction of students’ argumentation skills about controversial topics by epistemological understanding. Learning and Instruction, 16, 492–509.

Muis, K., Bendixen, L. D., and Haerle, F. C. (2006). Domain­generality and domain­specificity in personal epistemology research: Philosophical and empirical reflections in the development of a theoretical framework. Edu-cational Psychological Review, 18, 3–54.

Murphy, P. K., Alexander, P. A., Greene, J. A., and Edwards, M. N. (2007). Epistemological threads in the fabric of conceptual change research. In S. Vosniadou et al. (Eds.), Reframing the conceptual change approach in learn-ing and instruction (105–22). Oxford: Elsevier.

Murphy, P. K. and Mason, L. (2006). Changing knowledge and beliefs. In P. A. Alexander and P. Winne (Eds.), Handbook of educational psychology (2nd edn., 305–24). Mahwah, NJ: Lawrence Erlbaum Associates.

Norris, S. P. and Phillips, L. M. (1994). Interpreting pragmatic meaning when reading popular reports of science. Journal of Research in Science Teaching, 31(9), 947–68.

Perry, W. G. (1970). Forms of intellectual and ethical development in the college years: A scheme. New York: Holt, Rinehart and Winston.

Pintrich, P. R. (1999). Motivational beliefs as resources for and constraints on conceptual change. In W. Schnotz et al. (Eds.), New perspectives on concep-tual change (33–50). Amsterdam, The Netherlands: Pergamon.

Pintrich, P. R., Marx, R. W., and Boyle, R. B. (1993). Beyond cold conceptual change: The role of motivational beliefs and classroom contextual factors in the process of conceptual change. Review of Educational Research, 63, 167–99.

Posner, G. J., Strike, K. A., Hewson, P. W., and Gertzog, W. A. (1982). Accom­modation of a scientific conception: Towards a theory of conceptual change. Science Education, 67, 489–508.

Qian, G. and Alverman, D. (1995). Role of epistemological beliefs and learned helplessness in secondary school students’ learning science concepts from text. Journal of Educational Psychology, 87, 282–92.

(2000). Relationship between epistemological beliefs and conceptual change learning. Reading Writing Q, 16, 5–30.

Schommer, M. (1990). Effects of beliefs about the nature of knowledge on comprehension. Journal of Educational Psychology, 82, 498–504.

Beliefs about knowledge and revision of knowledge290

Schommer­Aikins, M. (2002). An evolving theoretical framework for an epis­temological belief system. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (103–18). Mahwah, NJ: Lawrence Erlbaum Associates.

Schommer, M., Mau, W., Brookhart, S., and Hutter, R. (2000). Understand­ing middle students’ beliefs about knowledge and learning using a multi­dimensional paradigm. Journal of Educational Research, 94, 120–7.

Schraw, G. (2001). Current themes and future directions in epistemological research: A commentary. Educational Psychology Review, 13, 451–65.

Schraw, G., Dunkle, M. E., and Bendixen, L. D. (1995). Cognitive processes in well­defined and ill­defined problem solving. Applied Cognitive Psychol-ogy, 9, 523–38.

Sinatra, G. M. (2005). The “warming trend” in conceptual change research: The legacy of Paul R. Pintrich. Educational Psychologist, 40, 107–15.

Sinatra, G. M. and Mason, L. (2008). Beyond knowledge: Learner characteris­tics influencing conceptual change (560–82). In S. Vosniadou (Ed.), Hand-book on conceptual change. Mahwah, NJ: Lawrence Erlbaum Associates.

Sinatra, G. M. and Pintrich, P. R. (2003). The role of intentions in conceptual change learning. In G. M. Sinatra and P. R. Pintrich (Eds.), Intentional conceptual change (1–18). Mahwah, NJ: Lawrence Erlbaum Associates.

Sinatra, G. M., Southerland, S. A., McConaughy, F., and Demastes, J. (2003). Intentions and beliefs in students’ understanding and acceptance of bio­logical evolution. Journal of Research in Science Teaching, 40, 510–28.

Smiley, P. A. and Dweck, C. S. (1994). Individual differences in achievement goals among young children. Child Development, 65, 1723–43.

Smith, C. L., Maclin, D., Houghton, C., and Hennessey, M. G. (2000). Sixth­grade students’ epistemologies of science: The impact of school science experiences on epistemological development. Cognition and Instruction, 18, 349–422.

Songer, N. B. and Linn, M. C. (1991). How do students’ views of science influ­ence knowledge integration? Journal of Research in Science Teaching, 28, 761–84.

Stanovich, K. E. (1999). Who is rational? Studies of individual differences in rea-soning. Mahwah, NJ: Lawrence Erlbaum Associates.

Stathopoulou, C. and Vosniadou, S. (2007a). Conceptual change in physics and physics­related epistemological beliefs: A relationship under scrutiny. In S. Vosniadou et al. (Eds.), Reframing the conceptual change approach in learning and instruction (145–65). Oxford: Elsevier.

(2007b). Exploring the relationship between physics­related epistemological beliefs and physics understanding. Contemporary Educational Psychology 32, 255–81.

Vosniadou, S. (2007). The conceptual change approach and its re­framing. In S. Vosniadou et al. (Eds.), Reframing the problem of conceptual change in learning and instruction (1–15). Oxford: Elsevier.

Vosniadou, S. and Brewer, W. F. (1987). Theories of knowledge restructuring in development. Review of Educational Research, 57, 51–67.

Lucia Mason 291

(1992). Mental models of the earth: A study of conceptual change in child­hood. Cognitive Psychology, 24, 535–85.

West, L. H. T. and Pines, A. L. (1985). Cognitive structure and conceptual change. Orlando, FL: Academic Press.

Windschitl, M. and Andre, T. (1998). Using computer simulations to enhance conceptual change: The roles of constructivist instruction and stu­dent epistemological beliefs. Journal of Research in Science Teaching, 35, 145–60.

Woolfolk Hoy, A., Pape, S. J., and Davis, H. (2006). Teacher knowledge and beliefs. In P. A. Alexander and P. H. Winne (Eds.), Handbook of educa-tional psychology (2nd edn., 715–37). Mahwah, NJ: Lawrence Erlbaum Associates.

292

10 The reflexive relation between students’ mathematics­related beliefs and the mathematics classroom culture

Erik De Corte, Peter Op ’t Eynde, Fien Depaepe, and Lieven VerschaffelUniversity of Leuven, Belgium

Introduction

Since the publication in 1983 of Schoenfeld’s seminal article “Beyond the purely cognitive: Belief systems, social cognition, and metacog­nitions as driving forces in intellectual performance,” a substantial amount of research has been carried out aiming at a better understand­ing of the nature and the structure of students’ mathematics­related beliefs and their relationships with student learning and performance (see Leder et al., 2002; Muis, 2004). Whereas this work paralleled the broader domain of inquiry on epistemological beliefs or personal epis­temology (Bendixen and Rule, 2004; Hofer and Pintrich, 1997, 2002; Hofer, 2004; Schommer­Aikins, 2002), it was mostly conducted sepa­rately from the latter strand of research.

According to the most common definition, epistemological beliefs are conceptions about the nature of knowledge and the nature of knowing (e.g., Hofer and Pintrich, 2002). Acknowledging that there are in the literature differences in conceptualization and in terminology, a widely accepted view defines the following dimensions of epistemological beliefs:

beliefs about the nature of knowledge involving the dimensions “cer­•tainty” (fixed or more fluid) and “simplicity” (discrete, concrete ver­sus complex, contextual) of knowledge;beliefs about the nature of knowing containing the dimensions •“source” (external versus self­constructed) and “justification” (crite­ria for knowledge claims, use of evidence) of knowledge (Hofer and Pintrich, 1997).

Although this view is supported by some empirical evidence (Hofer, 2000), it is obvious that the definition of epistemological beliefs remains

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 293

controversial among scholars in the field. Two major issues of debate in the literature on personal epistemology that are also relevant for the study of mathematics­related beliefs are: (1) whether beliefs about learning (and teaching) should be considered as epistemological (Hofer, 2002; Pintrich, 2002; Schommer­Aikins, 2002), and (2) whether epis­temological beliefs are domain­general or domain­specific (e.g., Buehl et al., 2002; Hofer, 2000).

According to Hofer and Pintrich (1997), beliefs about learning and teaching should not be included as part of personal epistemology, the core of which is constituted by the four dimensions mentioned above. And although they admit the importance of beliefs about learning, they rather consider them as developmental precursors of these core aspects of epistemology: children acquire a good amount of experience and knowledge about learning before their epistemological thinking fully develops. In contrast, Schommer explicitly includes beliefs about learn­ing, especially about the speed of learning and the control of learning, in her conception of personal epistemology (Schommer, 1994; Schom­mer­Aikins, 2002). As argued by Pintrich (2002), the empirical evi­dence in this regard is very mixed: some factor analytic studies show that beliefs about learning are separable dimensions from beliefs about knowledge and knowing, but other studies using interviews suggest a narrow relationship between beliefs about knowledge and about such aspects as learning, instruction, and intelligence. These divergences in theoretical conceptions and empirical outcomes certainly call for con­tinued inquiry. As we have shown elsewhere (Op ’t Eynde et al., 2002), in contrast to researchers studying personal epistemology in general, scholars who have focused on the nature and the role of mathematics­related beliefs have mostly included beliefs about learning and teach­ing, in addition to beliefs about mathematical knowledge and knowing, in their studies (e.g., Kloosterman, 1996; McLeod, 1992; Pekhonen, 1995; Underhill, 1988). This is not at all surprising; indeed, research results show that beliefs about learning are related to a variety of aspects of learning, such as perseverance when confronted with a difficult task, and planned thinking time (Schommer­Aikins, 2002).

Whether epistemological beliefs are domain­general or domain­spe­cific is also still in dispute (Op ’t Eynde et al., 2006), although there seems to be a growing consensus that to some degree epistemological thinking is domain­specific (Muis, 2004; Muis et al., 2006; Pintrich, 2002). Based on an analysis of the literature, Buehl and Alexander (2001) have suggested that epistemological beliefs are multidimen­sional, but also multilayered, involving three nested levels: general epis­temological beliefs, academic knowledge beliefs, and domain­specific

Mathematics­related beliefs and the classroom culture294

beliefs. Muis (2004) derives the following conclusion from her review of the literature on personal epistemology and mathematics: “The small amount of research that has been conducted in this line of inquiry has predominantly found support for the domain­specific hypothesis” (p. 346).

The subjects in most of the available investigations on this issue were students in higher education (e.g., Buehl et al., 2002; Hofer, 2000; Paulsen and Wells, 1998). One interview study involving sixty fifth­graders showed that their beliefs differed between mathematics and social studies. For instance, the students believed that they could learn social studies by themselves, but that they need the help of the teacher in mathematics (Stodolsky et al., 1991). Schoenfeld (1989) observed related differences in secondary school students: they considered math­ematical ability as innate, but believed that ability in English and social studies can be developed. Interestingly, the researchers of both stud­ies attribute these disparities in beliefs to differences in the teaching and classroom context between curriculum domains. One of our own studies with 365 junior high school students (aged fourteen) strongly suggests that beliefs relating to mathematics are domain­ and context­specific (Op ’t Eynde et al., 2006).

Against the background of the preceding discussion, this chapter aims at unravelling – based on the now available research – the com­plex relationship between students’ mathematics­related beliefs and the classroom culture.

Mathematics-related beliefs defined and framed in a mathematical disposition

We adopt here a broad view of students’ mathematics­related beliefs as implicitly or explicitly held subjective conceptions about mathematics education (involving beliefs about the domain of mathematics, about math learning and problem­solving, and about math teaching in gen­eral), about themselves as learners and knowers of mathematics (includ­ing self­efficacy, control, task­value, and goal­orientation beliefs), and about their mathematics class context, especially the classroom norms (De Corte, et al., 2002; Op ’t Eynde et al., 2002). As will be shown in the next section, these beliefs have – in close interaction with each other and with students’ prior knowledge – a strong impact on their math­ematical learning and problem­solving activities.

Fostering in students the acquisition of positive or availing mathematics­ related beliefs is at present regarded as an important goal of mathematics education. For instance, in the volume Adding it

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 295

up: Helping children learn mathematics, published by the US National Research Council (2001), one of the five strands of mathematical pro­ficiency is a “productive disposition” defined as a “habitual inclination to see mathematics as sensible, useful, and worthwhile, coupled with a belief in diligence and one’s own efficacy” (p. 5). This strand is con­sidered as a major component of what is called a mathematical disposi-tion, as illustrated by the following quote from the 1989 Curriculum and evaluation standards for school mathematics of the US National Council of Teachers of Mathematics (1989):

Learning mathematics extends beyond learning concepts, procedures, and their applications. It also includes developing a disposition toward mathemat­ics and seeing mathematics as a powerful way for looking at situations. Dis­position refers not simply to attitudes but to a tendency to think and to act in positive ways. (p. 230)

There is currently a broad consensus among scholars in the field of mathematics education (see, e.g., Baroody and Dowker, 2003; De Corte and Verschaffel, 2006; National Council of Teachers of Mathemat­ics, 1989, 2000; National Research Council, 2001; Schoenfeld, 1992) that building up and mastering a mathematical disposition requires the acquisition of five categories of cognitive, affective, and conative components, albeit that each distinct author uses a somewhat different terminology:

(1) A well­organized and flexibly accessible domain­specific knowledge base involving the facts, symbols, algorithms, concepts, and rules that constitute the contents of mathematics as a subject­matter field.

(2) Heuristics methods, i.e., search strategies for problem­solving which do not guarantee but significantly increase the probability of finding the correct solution because they induce a systematic approach to the task. Examples of heuristics include decomposing a problem into subgoals and making a graphic representation of a problem.

(3) Meta­knowledge, which involves knowledge about one’s cognitive functioning (e.g., knowing that one’s cognitive potential can be developed and improved through learning and effort), on the one hand, and knowledge about one’s motivation and emotions that can be used to deliberately improve volitional efficiency (e.g., becom­ing aware of one’s fear of failure when confronted with a complex mathematical task or problem), on the other hand.

(4) Positive mathematics­related beliefs, which include the implicitly and explicitly held subjective conceptions about mathematics

Mathematics­related beliefs and the classroom culture296

education, the self as a learner of mathematics, and the social context of the mathematics classroom.

(5) Self­regulatory skills, which embrace skills relating to the self­ regulation of one’s cognitive processes, such as meta­cognitive skills or cognitive self­regulation (e.g., planning and monitoring one’s problem­solving processes), on the one hand, and skills for regulat­ing one’s volitional processes/activities, such as meta­volitional skills or volitional self­regulation (e.g., keeping up one’s attention and motivation to solve a given problem), on the other hand.

It is important to stress here that these five components or strands of a mathematical disposition are interwoven, and, therefore, need to be acquired integratively. As argued in the above­mentioned report of the National Research Council (2001), a major challenge for mathematics education from preschool to grade eight is to make sure that children are making progress along every strand of the mathematical disposi­tion. In other words, fostering positive mathematics­related beliefs has to be pursued from the outset in the mathematics classroom, simulta­neously with the pursuit of the other strands. In the next section we address one of the three categories of math­related beliefs mentioned above, namely beliefs about mathematics education – especially beliefs about mathematics as a domain and beliefs about mathematics learning and problem­solving. In line with the scope of this book, we will focus on primary and secondary school students.

Students’ beliefs about mathematics and mathematics learning and problem-solving

Over the past two decades students’ beliefs about mathematics education at all levels of education have been studied using mainly questionnaires and interviews. The overall picture that emerges from reviews of the literature is that elementary and secondary school students, but even graduate students in mathematics, hold non­availing beliefs about mathematics and mathematics learning (De Corte et al., 2002; Muis, 2004). This common view about mathematics has been well character­ized by Lampert (1990) as follows: “Mathematics is associated with certainty, and with being able to give quickly the correct answer; doing mathematics corresponds to following rules prescribed by the teacher; knowing mathematics means being able to recall and use the correct rule when asked by the teacher; and an answer to a mathematical ques­tion or problem becomes true when it is approved by the authority of the teacher” (p. 31). In other words, students’ epistemological beliefs

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 297

about the nature of mathematical knowledge as certain and fixed in terms of one of the dimensions of Hofer and Pintrich’s theoretical framework (1997), and their beliefs about knowing and learning math­ematics whereby the source of new mathematical knowledge and its justification are – again in Hofer and Pintrich’s (1997) terms – located externally, do certainly not converge with the productive disposition as defined in the aforementioned National Research Council (2001) report. We will briefly document this here with some studies that yield additional confirmation for the outcomes reported in the reviews of the literature referred to above.

In a pertinent study of students’ beliefs about mathematics as a domain, Picker and Berry (2000) asked 476 twelve­ to thirteen­ year­olds from five countries (Finland, Romania, Sweden, United Kingdom, and the United States) to make a drawing of a mathematician and to comment on it in writing. Two distinct categories of drawings were obtained: some students depicted a mathematician who was clearly not a teacher, but others drew an image of a mathematician as a teacher. One major com­mon theme among the drawings and comments in the five countries is mathematics as coercion: the gist of the drawings of many of the students was indeed that of powerless little children confronted with a mathemati­cian depicted as authoritarian and threatening.

A second similarity that emerged from the data was the foolish math-ematician: mathematicians were often portrayed as lacking common sense, fashion sense, etc. According to the authors, this way of depict­ing a mathematician often also refers to what the students perceive as an unfair balance of power.

As it is plausible to assume that students’ drawings and the accom­panying comments reflect their beliefs about mathematics, it is obvious that they do not perceive this domain as attractive, interesting, and engaging. The major theme of mathematics as coercion that the draw­ings reveal is in accordance with several aspects mentioned above of the common (epistemological) beliefs about mathematics that derive from survey and interview studies: doing math is following the rules described by the teacher, knowing math is being able to recall and use the correct rule when asked by the teacher, and an answer to a problem becomes true when it is approved by the teacher. Interestingly, those domain­specific aspects of students’ beliefs converge nicely with the two dimensions of epistemological beliefs about the nature of knowing distinguished by Hofer and Pintrich (1997), namely source and justi­fication of knowledge. Indeed, the teacher is perceived as the external source of knowledge, and he/she also acts as the authority who justifies students’ answers.

Mathematics­related beliefs and the classroom culture298

With respect to learning, a strong belief that has been documented by many studies includes that students have the idea that mathematics is mainly a matter of memorization of facts, rules, and procedures (Muis, 2004). A recent interview study by Kloosterman (2002) with fifty­six high school students from grade nine to twelve provides additional evidence in this respect. Starting from the assumption that beliefs have a strong impact on students’ motivation and drawing on con­temporary relevant psychological theories, Kloosterman developed an extensive interview instrument involving fifty­one questions for assessing students’ mathematics­related beliefs. The questions cover a wide range of pertinent issues: feelings about school in general, non­school influences on motivation for mathematics, self­confidence, ability in mathematics, goal orientation, study habits, mathematics content, assessment practices, and students’ expectations of teachers. Major findings of this in­depth study are that students held the episte­mological beliefs that the nature of mathematics is procedural rather than conceptual, and that memorization and the ability to memorize procedures is crucial to be successful in mathematics. However, many students who stressed the importance of memorization claimed at the same time that people who are not good at memorizing can still do well in math if they work hard. An additional interesting outcome of this study is that many students had difficulty in answering the ques­tion about the nature of mathematics. According to Kloosterman, this seems to be an issue that students do not think about. He argues rightly that in view of successfully implementing innovative curricula students’ perceptions and beliefs about mathematics should be made more explicit.

A major aspect of students’ epistemology with respect to mathemat­ics is whether they see math as a powerful tool for modelling situations and objects in the real world. In this regard an important non­availing belief that has been reported in the literature is that the mathematics learned in school has little or nothing to do with the real world (Schoenfeld, 1992). Well­known work that illustrates this belief is the studies of street mathematics and school mathematics by Nunes et al. (1993) in Recife, Brazil. For instance, in one study Nunes et al. observed that young street vendors (nine­ to fifteen­year­olds) performed very well on prob­lems in the street­vending context (such as selling coconuts), but less well on isomorphic school mathematics tasks. In addition, they found that in the street­vending situation the children solved the problems using informal mathematical reasoning and calculation processes that differ considerably from the formal, school­prescribed procedures they tried to use with much less success on the textbook problems. These

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 299

findings show in a rather dramatic, albeit indirect way, the gap that can exist in children’s experience and beliefs between the world of school mathematics and the reality of everyday life.

This gap is also revealed in a more direct way in Boaler’s (1997; see also Boaler, 1998) classic in­depth study of two secondary schools in England, in which mathematics was taught according to strikingly different pedagogical approaches, namely traditional, textbook­based demonstrations and practice methods versus open­ended, project­based methods of teaching. For most students in the traditional school the mathematics they learned had no links with the real world as is illustrated by the following example of a student response:

Question: When you use maths out of school, does it feel different to using it in school or does it feel the same?

Answer: Well, when I’m out of school, the maths from here is nothing to do with it to tell the truth.

Additional question: What do you mean?Answer: Well, it’s nothing to do with this place, most of the things

we’ve learned in school we would never use anywhere. (p. 98)

Very suggestive of the belief in many students that real­world knowl­edge is irrelevant in the mathematics classroom is the work on the so­called phenomenon of suspension of sense-making (Schoenfeld, 1991). The most salient instance of this phenomenon in children’s problem­solving was already reported by French researchers in 1980 (Institut de Recherche sur l’Enseignement des Mathématiques de Grenoble, 1980). They administered to a group of first­ and second­graders the following absurd problem: “There are 26 sheep and 10 goats on a ship. How old is the captain?” It turned out that the large majority of the children produced a numerical answer (mostly 36) without any apparent awareness of the meaninglessness of the problem. Similar results were obtained in Germany (Radatz, 1983) and Switzerland (Reusser, 1986) with related problems. But the phenomenon also showed up in the US. The oft­cited example comes from the Third national assessment of educational progress with a sample of thirteen­year­olds (Carpenter et al., 1983): “An army bus holds 36 soldiers. If 1,128 soldiers are being bussed to their training site, how many buses are needed?” Of the over 70 percent of pupils who correctly carried out the division 1,128 by 36 obtaining the quotient 31 and remainder 12, only 23 percent gave “32 buses” as the answer; 19 percent gave the answer “31 buses,” and another 29 percent answered “31, remainder 12.” In all the above examples, pupils seem to be affected by the belief that real­world knowledge is irrelevant when solving mathematical

Mathematics­related beliefs and the classroom culture300

word problems, resulting in non­realistic mathematical modelling and problem­solving.

Using similar word problems under largely the same testing condi­tions, this phenomenon has been very extensively studied and repli­cated independently with pupils in the age range of nine to fourteen years during the 1990s, initially in several European countries (Bel­gium, Germany, Northern Ireland, and Switzerland), but also in other parts of the world (Japan and Venezuela) (for an extensive review, see Verschaffel et al., 2000; see also Greer et al., 2002). In the basic study (Verschaffel et al., 1994) a paper­and­pencil test consisting of ten pairs of problems was administered collectively to a group of seventy­five fifth graders (ten to eleven­year­old boys and girls). Each pair of prob­lems consisted of a standard problem, i.e., a problem that can be solved by the straightforward application of one or more arithmetic operations with the given numbers (e.g., “Steve bought 5 planks of 2 meters each. How many planks of 1 meter can he saw out of these planks?”), and a parallel problem in which the mathematical modelling assumptions are problematic, at least if one seriously takes into account the real­ities of the context called up by the problem statement (e.g., “Steve bought 4 planks of 2.5 meters each. How many planks of 1 meter can he saw out of these planks?”). An analysis of the pupils’ reactions to the problematic tasks yielded an alarmingly small number of realistic responses or comments based on the activation of real­world knowledge (responding to the problem about the 2.5 meter planks with 8 instead of 10). Indeed, only 17 percent of all the reactions to these ten problems could be considered as realistic, either because the realistic answer was given, or the non­realistic answer was accompanied by a realistic com­ment (e.g., with respect to the planks problem some pupils gave the answer 10, but added that Steve would have to glue together the four remaining pieces of .5 meters two by two). The fact that the studies mentioned above yielded very similar findings worldwide, justifies the conclusion that children’s belief that real­world knowledge is irrelevant when solving word problems in the mathematics classroom represents a very robust research result.

Additional studies in our center (De Bock et al., 2002; De Corte et al., 1999), but also by other European researchers (Greer and Verschaf­fel, 1997) have shown that this misbelief about the role of real­world knowledge during word problem­solving is very strong and resistant to change. For instance, in several investigations small interventions were inserted in the experimental setting aimed at sensitizing and stimulat­ing students to consider aspects of reality. Examples include alerting students at the onset of a written test that several problems are difficult

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 301

or impossible to solve because of unclarities in the problem statement; in interview studies a cognitive conflict was elicited by confronting students who gave a non­realistic response with the written notes of a fictitious classmate who had answered the same problem in a realistic manner. The results of these studies show that these kinds of interven­tions do not produce significant positive effects on student perform­ance (for an overview, see Verschaffel et al., 2000).

Research has thus yielded convincing evidence that students pre­dominantly hold beliefs about mathematics and about mathematics learning and problem­solving that are naïve and/or incorrect. It is plau­sible to assume that such beliefs have a negative and inhibitory impact on their learning and performance in mathematics. And, indeed, there is evidence supporting this hypothesis (for a detailed review, see Muis, 2004). For instance, in a survey study in the US with 230 students in grades ten through twelve, Schoenfeld (1989) found that those who obtained better results in math had more positive beliefs: they were less likely to perceive mathematics as a matter of memorization and less likely to believe that ability to memorize is crucial to succeed in math. In an investigation in Germany involving a large sample of over 2,000 upper secondary level students, Köller (2001) observed that those who believed that mathematical knowledge is certain and is constituted of isolated facts performed less well than students who perceived math knowledge as dynamic and interrelated. Of course, it is very likely that the relationship between beliefs and performance is reciprocal, in the sense that, for instance, again and again obtaining weak results on mathematics tests can have a negative impact on students’ beliefs. It is interesting to point out also here that the beliefs aspects identified by Köller (2001) in low performing upper secondary students, namely that knowledge is certain and constituted of isolated facts, converge with the two dimensions of epistemological beliefs about the nature of knowl­edge distinguished by Hofer and Pintrich (1997), respectively certainty and simplicity of knowledge.

However, it has also been argued that students’ non­availing beliefs do not necessarily have a negative impact on their mathematics achieve­ment, but are to the contrary rather functional in dealing with the stereotyped nature of the mathematical problems that prevail in the instructional as well as the evaluative settings of traditional mathemat­ics classrooms. Such problems do not at all stimulate students to invoke and apply their real­world knowledge (Reusser and Stebler, 1997; Ver­schaffel et al., 2000). This claim is – albeit indirectly – supported by the outcomes of another study by Schoenfeld (1988) in which he observed that students who did well on a standardized mathematics achievement

Mathematics­related beliefs and the classroom culture302

test held nevertheless naïve beliefs about problem­solving, such as “math problems should be solved in five minutes at the most” (p. xx).

All these findings certainly raise the important question of how do those negative beliefs originate and develop in students. In this respect there has evolved a broad consensus in the research community that those beliefs are largely induced by current classroom practices (e.g., Bendixen and Rule, 2004; Greeno, 1991; Lampert, 1990; Muis et al., 2006; Schoenfeld, 1988; Verschaffel et al., 2000), without denying, however, the impact of the wider context of the educational system as well as of culture and society (Greer et al., 2002). In this regard, Muis (2004) concludes her analysis of the literature on the development of mathematics­related beliefs as follows:

one plausible hypothesis is that formal education plays an important role in the development of students’ beliefs about the nature of mathematical knowledge and learning. (p. 339)

For a long time evidence for this hypothesis was scarce, albeit that anecdotal observations and a few case studies pointed in that direc­tion (see Verschaffel et al., 2000). Although in the context of the avail­able research the question about the impact of the classroom context on mathematics­related beliefs was often raised (see, e.g., Buehl et al., 2002; Hofer, 2001), those beliefs were mostly looked at from an individual learner perspective, reflecting the typical conception of mathematics as being a solitary activity (Seeger et al., 1998a). However, during the 1990s, under the influence of the socio­cultural and socio­ constructivist approaches to the study of learning, the importance of the social context of learning for the development of student beliefs about mathematics and mathematics learning was more and more stressed, especially the role of the classroom culture in general and the classroom norms and practices in particular (e.g., Cobb and Yackel, 1998; Seeger et al., 1998b).

A socio-constructivist perspective on the reflexive relationship between classroom culture and mathematics-related beliefs

In view of unravelling the relationships between classroom norms and practices on the one hand, and beliefs on the other hand, it is important to realize from the outset that these norms and practices are them­selves shaped through the interactions between teachers and students, which are, in turn, co­determined by their respective mathematics­ related beliefs. Understanding the way in which students’ and teachers’

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 303

beliefs influence learning and teaching in the mathematics classroom then involves the full recognition of the complex and reflexive nature of the relations between the classroom culture and those beliefs (see also Seeger et al., 1998a).

A prominent and influential socio­constructivist approach to the study of the culture of the classroom is undoubtedly Cobb’s interpretive framework for analyzing the individual and social dimensions of class­room activities. Cobb and his associates (Cobb and Yackel, 1998; Cobb et al., 2001; McClain and Cobb, 2001; Yackel and Cobb, 1996) describe the classroom micro­culture in terms of classroom norms and prac­tices on the one hand, and teachers’ and students’ beliefs, conceptions, and activities on the other hand. These clusters of concepts represent, respectively, the social and psychological perspectives underlying socio­constructivism: the social perspective refers to ways of acting, reason­ing, and arguing that are normative in a classroom community, while the psychological perspective is concerned with the nature of individual students’ reasoning, or their particular ways of participating in commu­nal activities (Cobb et al., 2001).

In the social perspective, a distinction is made between classroom social norms, socio­mathematical norms, and classroom mathematical practices. Classroom social norms are “characteristics of the classroom community and document regularities in classroom activity that are jointly established by the teacher and students” (Cobb et al., 2001, pp. 122–3). For instance, in the classroom students have to listen to each other and they have to raise their hand before answering. Whereas social norms apply to any subject­matter domain of the curriculum, socio­mathematical norms are domain­specific in the sense that they bear on normative aspects of students’ mathematical activities and dis­cussions (Cobb et al., 2001). Examples include: what is considered as a mathematically correct answer, as a mathematically different solution method, as a mathematically appropriate explanation or justification. Finally, classroom mathematical practices are defined as the taken­as­shared patterns of activity established by the classroom community (Cobb and Yackel, 1998). Specific procedures that the teacher and the students agreed upon to solve certain mathematical problems are typi­cal examples of classroom mathematical practices.

The psychological perspective of Cobb and associates’ framework also consists of three categories and refers mainly to students’ beliefs about mathematics (Cobb and Yackel, 1998), namely: (1) beliefs about the personal role and the role of others in the mathematics classroom and about the general nature of mathematical activity; (2) specifically mathematical beliefs and values; and (3) mathematical conceptions

Mathematics­related beliefs and the classroom culture304

and activity. The first category refers to “students’ interpretations of their own and others’ activity” (Cobb and Yackel, 1998, p. 168). Spe­cifically mathematical beliefs enable students “to act as increasingly autonomous members of the classroom mathematical community as they participate in the negotiation of sociomathematical norms” (Cobb et al., 2001, p. 124). Mathematical conceptions and activities are seen as acts “of individual learning in which a student reorganizes his or her mathematical reasoning” (Cobb et al., 2001, p. 125).

The psychological and the social perspectives represent two differ­ent, but complementary ways of looking at and making sense of what is going on in classrooms. Each perspective constitutes the background from which the other approach can be better understood. Conse­quently, both perspectives, and thus also their three subcategories, are reflexively related:

normative activities of the classroom community (social perspective) emerge and are continually regenerated by the teacher and students as they interpret and respond to each other’s actions (psychological perspective). Conversely, the teacher’s and students’ interpretations and actions in the classroom ( psychological perspective) do not exist except as acts of participation in communal classroom practices. (Cobb et al., 2001, p. 122)

In a more recent contribution, Cobb et al. (2006) have elaborated their theoretical perspective using the notion of identity, a concept to which more and more attention has been paid recently in mathematics education research (e.g., Boaler, 1999; Boaler and Greeno, 2000; Nasir, 2002). Defined in Webster’s third new international dictionary (Gove, 1993) as the “unity and persistence of personality” (p. 1123), identity has traditionally been conceived of as a construct that refers to the indi­vidual. But, more recently, the concept has been defined in terms of the relationship and interaction between the individual and the social context (see, e.g., Wenger, 1998). As argued by Boaler (2000):

Such a construct [of identity] brings together sociological and psychological theories of learning, capturing the relationship between individual knowledge and beliefs and the broader communities in which knowledge and beliefs are developed and used. (p. 11)

In other words, the notion of identity is seen as a bridging construct between the reciprocal individual and social dimensions of classroom learning.

In the next two sections we will review brief ly two kinds of recent empirical studies that fit within this socio­constructive perspective on the relationship between the classroom culture and students’

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 305

mathematics­related beliefs. Although not each of them refers explicitly to Cobb’s theoretical framework, they all attempt to con­tribute to unravel the complex interactions among the social and the psychological perspective: (1) ascertaining or descriptive stud­ies focusing on systematic analyses of existing mathematics class­room practices (Boaler and Greeno, 2000; Depaepe et al., 2007); and (2) intervention studies in which a learning environment was created aimed at fostering more positive mathematics­related beliefs in students (Mason and Scrivani, 2004; McClain and Cobb, 2001; Verschaffel et al., 1999).

Ascertaining studies

A study by Boaler and Greeno (2000)

Research method. Boaler and Greeno interviewed forty­eight high school students, eight in each of six classes of different schools. All students had chosen to take an advanced level calculus class, and could thus be considered as good math students. The six experienced teachers of these classes each had a good reputation. The semi­structured inter­views were done with single­sex pairs of students and focused on the nature of the math lessons, lessons they liked and disliked, the extent of discussion in the mathematics classroom, and their mathematical confidence. The semi­structured nature of the interviews allowed for the collection of rich, in­depth data that were processed using an open coding system.

Results. The classroom culture in four of the six classrooms was described by the students as structured, individualized, and ritualized; in contrast, the culture in the other two classes was characterized as relational, communicative, and connected. The authors label these two so­called ecologies of participation, respectively, as didactic teaching and discussion­based teaching, and they observed a clear relationship between the two ecologies or classroom cultures on the one hand, and students’ mathematics­related beliefs on the other hand.

Independent of gender, confidence level, and attainment, students in the didactic­teaching classes perceived mathematics as a collection of conceptually opaque procedures, and their own role in the math class as passive absorbers of those procedures through memorization. Some of them held the beliefs that for math tasks there is always one correct answer, and that in mathematics – as opposed to other subjects – think­ing is not required. The following extract from an interview is illustra­tive: “There’s definitely a right answer to it. The other subjects like

Mathematics­related beliefs and the classroom culture306

English and stuff that really have no right answer, so I have to think about it” (Boaler and Greeno, 2000, p. 179).

According to the authors this ecology of didactic teaching is grounded in an epistemology of received knowing. The latter concept, borrowed from the work of Belenky et al. (1986) on epistemological beliefs, means that a person considers knowledge as derived from an external authori­tative source outside the self. This classroom culture apparently also had a strong impact on students’ developing identities that was in line with a procedure­oriented and memorization­based approach to mathematics learning. When asked about how to be successful in their math class, they did not stress ability, but the willingness of accepting the role of receiver of knowledge involving compliance, perseverance, obedience, and frustration. A remarkable finding is that the students in these four classes who reported that they liked mathematics did so exactly because they held the epistemological beliefs that there were in this subject only right and wrong answers and that there was no need to be creative or reflective.

About half of the – otherwise successful – students in the didactic classes rejected mathematics. Interestingly is that also here the reasons they mentioned were primarily related to the development of their iden­tities. They did not refer to inability for or disliking of mathematics, but to the lack of opportunity for creativity, interpretation, and agency, as illustrated in the following extract (Boaler and Greeno, 2000, p. 186):

i n t: Why wouldn’t you major in math?s t: I think I’m a more creative person. I can do it and I can understand it but

it’s not something I could do for the rest of my life and I think if I had a job I’d like one that let me be a little more creative.

i n t: Math isn’t creative …?s t: No.

In contrast with the didactic classes, students who were immersed in discussion­based teaching perceived mathematics as a domain of inquiry open to exploration, negotiation, and discussion, and their role as active learners focused on understanding connected mathematical knowledge and relationships.

According to Boaler and Greeno, discussion­based teaching ref lects what Belenky et al. (1986) called an epistemology of connected knowing: an individual constructs knowledge actively in interaction with others. The interview data showed also that the students who were immersed in this alternative classroom culture developed dif­ferent patterns of identity than those in the didactic classes. They reported more positive identifications with mathematics, and the reasons were mostly that math provided opportunities for thinking

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 307

and for acquiring connected understanding, and that they were given more agency.

Eight out of the ten students in the discussion­based classes who were asked about it, mentioned that they would take other mathemat­ics courses in the future. All but one of the sixteen students in these classes said that they enjoyed mathematics. Interestingly, seventeen of all forty­eight good math students who were interviewed in this study reported that they hated or disliked math; and sixteen of those belonged to the didactic classes (which is half of the students from those classes).

Conclusion. The results of this study show that in this sample of high school students there was a clear relationship between the classroom culture on the one hand, and students’ mathematics­related beliefs and developing identities on the other hand. As such, this investigation sup­ports the hypothesis mentioned above that Muis (2004) derived from her literature review, namely that formal education has a strong impact on students’ (epistemological) beliefs about the nature of mathemati­cal knowledge and learning (even though the study does not yet pro­vide a detailed description of the explicit and/or implicit interactional processes through which the educational practices and culture exactly impact students’ beliefs). But it is also plausible that – reciprocally – stu­dents’ beliefs that transpire in their behaviour during the mathematics lessons influence the evolving classroom culture and practices. This would be the case when, in a mathematics class that is dominated by an epistemology of received knowing, one or a few students who hold (due to participation in previous mathematics classes or to out­of­school experiences) an epistemology of connected knowing and who behave accordingly, begin to question – explicitly or implicitly – the norms and practices in that mathematics class, which may result in the develop­ment of a new classroom culture and practices that are more in line with the minority’s initial (epistemological) beliefs.

A study by Depaepe, De Corte, and Verschaffel (2007)

Objective. By analyzing teaching practices that start from a reform­based mathematics textbook, we aimed in this study at documenting the relationship between the social (i.e., the classroom culture) and the individual (i.e., teachers’ and students’ mathematics­related beliefs) perspectives in problem­solving lessons. An additional objective was to see whether textbook lessons that are explicitly based on an innova­tive approach to teaching problem­solving are implemented with a high degree of fidelity at the classroom level.

Mathematics­related beliefs and the classroom culture308

Research method. In ten sixth­grade classrooms in which the reform­based textbook Eurobasis was used, we videotaped the same two problem­solving lessons taught by the regular classroom teachers. To document the social perspective we developed a coding system that addresses three important aspects of the classroom culture and prac­tices which are considered to influence students’ mathematics­related beliefs and their mathematical modelling competence: (1) the teach­ers’ focus – as evidenced by his/her speech acts – on each of fourteen heuristic and metacognitive skills on the one hand, and on the explicit negotiation of ten different norms on the other; (2) the use of each of eight instructional techniques and four classroom organization forms; and (3) the realistic nature and the degree of complexity of the prob­lems. Three instruments were administered to map the individual per­spective: teachers’ and students’ beliefs were measured by means of questionnaires in the format of a five­point Likert­scale (with fifteen and twenty­one statements, respectively), and students’ (non­)realistic mathematical modelling was assessed by a paper­and­pencil test con­taining ten non­routine problems.

Results. With respect to the social perspective the analysis of the teachers’ speech acts showed that some heuristic and meta­cognitive skills were addressed quite frequently (e.g., distinguish relevant from irrelevant data, make a scheme/table), but others almost never (e.g., simplify the numbers in the problem, contextualize a calculation); and only in very few cases did a teacher explicitly emphasize why it is impor­tant to use a heuristic or meta­cognitive skill. Throughout the lessons hardly any attention was paid by the teachers to the systematic and intentional explicitation and negotiation of norms. Apart from model­ling, scaffolding, and exploration, the other instructional techniques from the cognitive apprenticeship model (Collins et al., 1989) were frequently employed. Notably in this respect is that besides (directive and non­directive) coaching, reflection, and especially articulation, were quite well used; both techniques are certainly relevant to foster students’ problem­solving competence. Whole­class instruction was the most frequently used classroom organization form. However, only three out of the ten teachers applied group work rather frequently. This is remarkable taking into account, on the one hand, that over the past fifteen to twenty years the mathematics education reform literature has stressed the importance of learning in small groups for the develop­ment of problem­solving skills in students (see, e.g., Good et al., 1992), and, on the other hand, that group work is a rather usual instructional technique in other domains of the curriculum. Whereas most of the problems used during the lessons had some relationship to children’s

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 309

experiential world, the percentage of complex tasks requiring the acti­vation of thinking skills was very low. It has to be added that on all aspects discussed above there were substantial differences among the ten teachers.

Teachers’ questionnaires revealed that in general they had positive beliefs toward (the teaching) of mathematical problem­solving. How­ever, three of the ten teachers still believed that a good problem has an exact solution. But in contrast to the teachers, the beliefs of the students were only moderately positive. Moreover, their scores on the test con­sisting of ten non­routine word problems were rather low, ranging from two to five out of ten (suggesting at the same time that there were also here significant differences among the teachers). Also, students’ scores on the test and on the beliefs questionnaire correlated significantly. The small sample of this study does not allow for generalization, but the latter findings suggest that attempts at improving problem­solving teaching by introducing a new approach to instruction does not easily result in the enhancement of students’ mathematics­related beliefs, nor in the increase of their competence in mathematical modelling.

Our attempt in this study at contributing to unravelling the relation­ship between the individual and the social perspective in mathematics problem­solving lessons has not gone very far. Only some provisional but promising trends were observed as illustrated by the following exam­ples. First, only one teacher frequently stressed the norm that problems can be solved in different ways. This was also the only class where all students expressed the belief that math problems can indeed be solved in different ways (59 percent totally agreed and 41 percent agreed with this statement); in all other classes the overall agreement with this state­ment was considerably lower. Second, the class that performed the worst on the problem­solving test had the least availing beliefs. Compared to her colleagues, the teacher of that class coached more directively. This more directive teacher also discussed only twice explicitly an appropri­ate norm for mathematical problem­solving, emphasized only twice the why, how, and when to use a particular heuristic, and never used group work.

Conclusion. This study clearly shows that introducing in a textbook an innovative approach to the teaching of mathematical problem­solving leads to a large variety in the quality of implementation of the underly­ing reform ideas, and does certainly not automatically result in a high­fidelity implementation of the new approach. But, more importantly, the investigation documents that our approach has not been very suc­cessful in advancing our deeper understanding of how different aspects of the classroom culture and practices influence and interact with

Mathematics­related beliefs and the classroom culture310

individual students’ learning activities and beliefs relating to mathe­matical problem­solving. In this respect some methodological remarks are in order. Our dataset covered only two lessons in each classroom, whereas there is no doubt that norms and beliefs develop interactively and gradually over time. Since we intentionally only coded skills and norms that were explicitly addressed by the teacher, our analysis is based exclusively on low­inference coding. But norms are also negotiated and develop implicitly, so that we could only partially grasp the ways of acting, reasoning, and arguing that were normative in the classrooms. Therefore, in order to be able to better document the implicitly as well as the explicitly negotiated classroom norms and their interaction with students’ learning and beliefs, we are conducting a more in­depth case study in the classrooms of two teachers who – although they use the same textbook – differ in their approach to teaching mathematical problem­solving. This study is spread over a period of seven months of one school year, and – as will be explained in the final section of this chapter – involves a broader range of research techniques.

Intervention studies

In this section we review three intervention studies, i.e., design or teaching experiments in which learning environments were created that embody a new classroom culture and aim at fostering more posi­tive mathematics­related beliefs in students.

A study by McClain and Cobb (2001)

Objective. Using the interpretive framework discussed above, Cobb and his colleagues have over the past ten years undertaken a series of class­room teaching experiments, or design experiments, in lower primary classrooms, varying in duration from just a few weeks to an entire school year. The aim of a classroom teaching experiment is to study students’ mathematics learning in novel learning environments designed in col­laboration with a teacher who becomes a member of the research team (Cobb, 2000). The present study of McClain and Cobb is an example of such an experiment in one first­grade class with eighteen students. The work aimed at analyzing how the teacher proactively fostered the establishment of certain socio­mathematical norms and, by doing so, simultaneously attempted to enhance children’s mathematics learning and the development of a mathematical disposition.

Research method. A variety of data was collected throughout the experiments that lasted one school year, and aimed at the development

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 311

of teaching activities that addressed key mathematical concepts that relate to mental calculation and estimation. Video­recordings of 103 lessons were made using mostly two cameras, one focused mainly on the teacher, but sometimes on individual children who explain their rea­soning and problem­solving; the other camera registered pupils while they are involved in discussions about a math task. Other data sources are: copies of students’ written work, field notes relating to the daily les­sons, reports of the daily and weekly planning and debriefing sessions of the researchers together with the teacher, the teacher’s diary, and video registrations of individual interviews with pupils. The method used to analyse those data is in line with the constant comparison method of Glaser and Strauss (1967) applied in ethnographic studies.

Results. We mention first that interview data collected throughout the school year showed that all students made substantial mathemati­cal progress. The findings with respect to the development of socio­ mathematical norms are based on the video­data of lessons during the first four months of the school year. McClain and Cobb (2001) have shown what first­grade teachers could do to evoke and sustain the development of socio­mathematical norms at the classroom level and mathe matics­related beliefs in individual children that are in line with the mathematical disposition advocated in current reform documents. The analyses related to the following norms: what counts as an accepta­ble explanation, what counts as a mathematically different solution, and what counts as an easy, simple, and efficient solution. As an illustration of the results we focus here on the norm mathematically different.

One task given to the children was to figure out how many chips there were shown on an overhead projector on which an arrangement of, for instance, five or seven chips was displayed for three or four sec­onds. The objective was to elicit reasoning about the task, and to initi­ate a shift in pupils’ methods from using counting to find the answer toward using more sophisticated strategies based on grouping of chips. The results reveal how, through discussions and interactions focused on the task, the “mathematical difference norm” developed in the class­room, as shown in the following example (McClain and Cobb, 2001, pp. 250–1).

The teacher showed seven chips arranged as horizontal rows of two, three, and two. Several students explained that they had counted, and Dan stated that he had seen six and one more.Next the teacher asked for solutions different from six and one.

t e ac h e r : Now, Jane, are you sure you have a different way, a different way not to count but to see it?

Mathematics­related beliefs and the classroom culture312

j a n e : Five plus two.t e ac h e r : Would you show us five plus two?j a n e: (Points to a group of five composed of the first two rows and the two

remaining chips in the last row.)t e ac h e r : Now, Jane, you really understood what I meant. (Carl raises his

hand.) Now, Carl, are you sure it’s not just a different way to count by one?

c a r l: You mean going across is not different?’

At this point, there is every indication that a basis for communication that would make productive discussions possible was emerging. For example, Jane judged that her solution was different when compared to those that had already been offered and, in doing so, contributed to the criteria for what counted as different. We would argue that the written notation that clarified Dan’s solu­tion made it possible for Jane to compare and contrast her solution method to his in order to determine if hers would be considered different. For this reason, the notation played a critical role in initiating shifts in the students’ ways of reasoning about the task. The goal or purpose of the task was no longer just determining the number of chips. The purpose now entailed identifying a dif­ferent way to group the chips to find the total. As a result, a different solution was one that was mathematically different from solution processes that had already been offered.

Later on during the class discussions the mathematical difference norm evolved into a renegotiation of the norm “what is a sophisticated solution?” Indeed, solutions based on the grouping of chips became seen not only as different from, but also as more sophisticated than, counting. It is plausible to assume that in parallel with the emergence of those socio­mathematical norms, students’ individual beliefs about mathematics and mathematics learning were influenced, and that this contributed to their acquisition of a mathematical disposition.

Conclusion. This study shows how socio­mathematical norms in the micro­culture of the classroom emerge, evolve, and further develop throughout interactions between teacher and students, and also how these norms regulate continued classroom discourse and contribute to the creation of learning opportunities. But, as Cobb (2000) con­cedes, due to the focus on the classroom as a community of learners this type of experiment is less appropriate for investigating and docu­menting mathematical learning of individual students, and, therefore, falls short of well documenting the psychological perspective of Cobb’s interpretive framework. Although it is likely that correlatively with the establishment of new norms embedded in the classroom practices the mathematics­related beliefs of individual students develop, those beliefs are not at all operationalized and assessed in the reports of this and other similar experiments, although it might not be too difficult to do so.

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 313

A study by Verschaffel, De Corte, Lasure, Van Vaerenbergh, Bogaerts, and Ratinckx (1999)

Objective. As a contribution to the implementation of the new standards for mathematics in primary education in the Flemish part of Belgium, Verschaffel et al. (1999) carried out a study aiming at the design and evaluation of a learning environment that is powerful in eliciting in upper primary school children the appropriate learning processes for acquiring the intended competence in mathematical problem­solving as well as positive mathematics­related beliefs.1

Research method. The intervention in the four participating experi­mental fifth­grade classes consisted of twenty lessons that were taught by the regular classroom teachers. The learning environment was fun­damentally changed in narrow cooperation with the four teachers with respect to the following components: the content of learning and teach­ing, the nature of the problems, the instructional techniques, and the classroom culture. First, in terms of content the learning environment focused on the acquisition by the students of an overall meta­cognitive strategy for solving mathematical application problems consisting of five stages: build a mental representation of the problem, decide how to solve the problem, execute the necessary calculations, interpret the out­come and formulate an answer, and verify the solution. Eight heuristic strategies, which are especially valuable in the first two stages, were embedded in the teaching of that strategy.

Second, a varied set of carefully designed realistic, complex, and open problems were used that differ substantially from the traditional textbook tasks. Third, a learning community was created through the application of a varied set of activating and interactive instructional techniques, especially small­group work and whole­class discussion. Throughout the whole lesson the teacher’s role was to encourage and scaffold pupils to engage in, and to reflect upon, the kinds of cognitive and meta­cognitive activities involved in the model of skilled problem­solving. These instructional supports were gradually faded out as pupils became more competent in and aware of their problem­solving activity, and thus took more responsibility for their own learning and problem­solving processes. Fourth, an innovative classroom culture was created through the establishment of new norms about learning and teaching problem­solving, and aiming at fostering positive mathematics­related beliefs. Typical aspects of this classroom culture are: (1) stimulating

1 Actually this design experiment was a major source of inspiration for the reform­based textbook Eurobasis used by the teachers who participated in the ascertaining study of Depaepe et al. (2007) described above.

Mathematics­related beliefs and the classroom culture314

pupils to articulate and reflect upon their solution strategies, beliefs, and feelings relating to mathematical problem­solving; (2) discussing what counts as a good problem, a good response, and a good solu­tion procedure (e.g., there are often different ways to solve a problem; for some problems a rough estimate is a better answer than an exact number); and (3) reconsidering the role of the teacher and the pupils in the mathematics classroom (e.g., the class as a whole will decide which of the generated solutions is the optimal one after an evaluation of the pros and cons of the different alternatives).

The effects of the learning environment were evaluated in an experi­ment with a pretest–posttest–retention test design with an experimen­tal group (four classes) and a comparable control group (eight classes), using thereby a wide variety of data­gathering and analysis techniques. Three parallel versions of a written test consisting of ten difficult non­routine problems were developed and used as pretest, posttest, and retention test, respectively. A Likert­type beliefs questionnaire relat­ing to mathematical word problem­solving and a standardized math­ematics achievement test were both applied as pretest and posttest. The beliefs questionnaire consists of two subscales based on factor analy­sis: a first subscale containing seven items deals with pupils’ pleasure and persistence in solving word problems (for instance: I like to solve word problems; difficult problems are my favourites), and a second subscale with fourteen items expressing a problem- and process-oriented view on word problem-solving (for instance: there is always only one solution to a word problem; listening to explanations of alternative solution paths by other pupils is a waste of time). In addition, in each of the four experi­mental classes the solution processes of three pairs of children for five non­routine problems were videotaped and analyzed before and after the intervention.

Results. According to the scores on the word problem pretest and the parallel posttest and retention test, the intervention had a significant and stable positive effect on the experimental students’ problem­solving performance. The questionnaire data indicated that the learning envi­ronment also had a significant, albeit small, positive impact on chil­dren’s pleasure and persistence in solving mathematics problems, and on their problem­ and process­oriented view of word problem­solving. The analysis of pupils’ written notes on their response sheets of the word problem test showed that the better results of the experimental children were paralleled by a strong increase in the spontaneous use of the heuristic strategies taught in the learning environment; this finding was confirmed by a qualitative analysis of videotapes of the problem­solving processes of three pairs of children from each experimental

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 315

class before and after the intervention. Finally, we found that not only the high and the medium ability pupils, but also those of low ability benefited significantly – albeit to a smaller degree – from the interven­tion in all aforementioned aspects.

Conclusion. The results of this study support the hypothesis that a substantially modified learning environment, combining a set of care­fully designed word problems with highly interactive teaching methods and the introduction of a new classroom culture, can lead to the crea­tion of a high­powered learning community which not only significantly boosts students’ cognitive and meta­cognitive competency in solving mathematical word problems, but also fosters their mathematics­related (epistemological) beliefs. However, the latter effect was quite small, and whereas this can probably be partially explained by the short duration of the intervention, the observational data in the experimental classes also indicated that the establishment of a new classroom culture in view of influencing students’ beliefs was not realized in a sufficiently system­atic and effective way. Below we review a recent study that was more successful in this respect.

A study by Mason and Scrivani (2004)

Objective. The study by Mason and Scrivani can to a large degree be considered as a replication of the preceding investigation by Verschaffel et al. (1999). The major objectives were to test the following hypoth­eses: (1) a novel instructional intervention that substantially differs from the traditional learning environment can have a positive impact on students’ mathematics­related beliefs; and (2) as beliefs are assumed to affect students’ problem­solving, the intervention will also have a positive impact on their performance. In addition the study intended to see whether the new learning environment had a favourable influ­ence on students’ perception of their own effort and understanding in mathematics.

Research method. Eighty­six students of four fifth­grade classes par­ticipated in this study. In two classes involving forty students, an inno­vative learning environment was implemented; the forty­six children of the other two classes were taught in the traditional way according to the regular mathematics curriculum. Both groups were randomly assigned to the two conditions.

The intervention consisted of twelve sessions of one and a half hours taught weekly over a three­month period by one of the researchers, an experienced math teacher. The learning environment had the fol­lowing characteristics. First of all, a more systematic attempt than in

Mathematics­related beliefs and the classroom culture316

the previous study was made to create an innovative classroom culture through the negotiation of new norms relating to the role of the stu­dents and the teacher, and concerning what counts as a good problem, a good solution method, and a good answer. The teacher encouraged the students to engage in the activities involved in the overall meta­cognitive strategy developed in the preceding study of Verschaffel et al. (2000), and to take progressively more agency for their problem­ solving activities. Second, the same activating and interactive instructional techniques as in the previous investigation were used, especially small­group work and whole­class discussion. Third, in order to elicit in stu­dents reflection about problems, they were confronted with a variety of non­routine tasks, which often required realistic modelling and/or could be solved in different ways and could have distinct solutions.

Based on instruments available in the literature, the authors devel­oped a Likert­type beliefs questionnaire with twenty­eight statements focusing on two of the three categories of mathematics­related beliefs distinguished by De Corte et al. (2002): (1) beliefs about mathematics, mathematics learning, and problem­solving; and (2) beliefs about the self as a learner of mathematics (perceived ability and possibility to develop ability). The questionnaire was administered before and after the intervention. Students’ performance was measured using two kinds of word problems as pretest and posttest: two usual and two unusual problems. The latter problems, aimed at the assessment of students’ sense­making, were substantially different from the common tasks in math classrooms: one was a realistic problem (a variant of the “army bus” problem mentioned above), and the other an unsolvable, absurd one (a variant of the “how old is the captain” task). Finally, students were asked before and after the intervention to evaluate on a five­point Likert scale their own effort in and understanding of mathematics.

Results. A Rasch model analysis confirmed the two beliefs catego­ries used as the basis for the construction of the questionnaire: beliefs about math and mathematical learning and problem­solving, and beliefs about the self as a learner of mathematics. The questionnaire data revealed that – whereas initially the two groups did not differ in terms of their math­related beliefs – those who were immersed in the innovative learning environment had afterwards significantly more positive beliefs in both categories compared to the traditional instruc­tion group. Interestingly, the effect of the intervention on students’ beliefs was much stronger than in the aforementioned study. The data also confirmed the second hypothesis: after the intervention the innovative learning environ ment group outperformed the traditional teaching group in both the usual and the unusual problems. Moreover,

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 317

performance on the unusual problems after the intervention correlated, albeit only moderately, with both beliefs categories: students with more availing beliefs performed better on the non­routine problems. Finally, the intervention also had a positive effect on students’ perception of effort and understanding in mathematics in the innovative learning environment group, but not in the traditional teaching group.

Conclusion. This study replicates a major finding of the previous inves­tigation, namely the significant positive effect of the learning environ­ment on students’ performance in mathematical problem­solving. But, more importantly, Mason and Scrivani have shown that by focusing in the intervention more systematically on the establishment of new socio­mathematical norms, substantial gains can also be achieved with respect to students’ beliefs; indeed, they seem to acquire more avail­ing beliefs about mathematics and mathematics learning and problem­solving, but as well about themselves as learners of mathematics.

Taking into account the small number of classes involved in the three teaching experiments reviewed, it would be premature to derive already very general conclusions concerning the appropriateness and the effec­tiveness of the instructional interventions involved in view of fostering in students in regular primary school classrooms positive mathematics­related beliefs. This is the more so considering that in the study by Ver­schaffel et al. (1999; see also Verschaffel et al., 2000) very substantial support was provided to the participating teachers in view of achiev­ing high­fidelity implementation of the learning environment. And, as shown in the ascertaining study of Depaepe et al. (2007) discussed above, such a high­fidelity implementation is not easily realized without such support (see also Remillard, 2005).

Conclusion, future research, and educational implications

Major results of past research

Research on epistemological beliefs in general and on mathematics­related beliefs in particular is a rather recent phenomenon, and had only reached cruising speed by the 1990s (Leder et al., 2002; Muis, 2004). Over this short period of time important results have already been achieved leading to a general consensus that beliefs about math­ematics as a domain, about mathematics learning and teaching, and about oneself as a learner of mathematics have, besides and in interac­tion with cognitive variables, an important impact on students’ learn­ing and performance in school mathematics. As shown in this chapter,

Mathematics­related beliefs and the classroom culture318

substantial progress has been realized in identifying a variety of (often unavailing) beliefs relating to mathematics in students; also, attempts have and are being made at developing comprehensive and integrative models of mathematics­related beliefs (e.g., Op ’t Eynde et al., 2002). Furthermore, there is already convincing evidence in the literature that beliefs and performance in mathematics correlate positively. Finally, intervention studies indicate that by immersing students in innova­tive learning environments it is possible to foster in them more positive mathematics­related beliefs. But, due to the multidimensional nature of these interventions, it is impossible to derive the relative contribution of the different aspects of the learning environment to the observed gains in students’ beliefs and performance. An assumption underly­ing this chapter evidently is that the social dimensions of classroom practices, especially the classroom culture, are of utmost importance in this regard. In an attempt to unravel the reflexive relationships between classroom culture and individual beliefs, we have used the socio­constructivist perspective as a framework for a selective discus­sion of studies that address those relations. As theoretical background for this discussion we have described in more detail Cobb’s (2000) interpretive framework that represents an influential elaboration of the socio­ constructivist perspective for analyzing individual and collective activity at the classroom level.

Although substantial progress has thus been made in analyzing and understanding students’ mathematics­related beliefs, some critical theo retical and methodological remarks have to be made with respect to several of the research aspects mentioned in the previous paragraph. Those comments point at the same time to directions and issues for continued inquiry.

Theoretical and methodological comments on past research

Although there is no doubt that Cobb’s (2000) interpretive frame­work offers a helpful set of constructs for studying and understand­ing classroom processes and practices, it also raises a series of critical comments.

First, a major aspect of Cobb’s framework is the distinction between social and socio­mathematical norms, the former being domain­ independent and the latter specifically related to activities in the math­ematics classroom. As has been argued by Mottier Lopez (2005; see also Mottier Lopez and Allal, 2007), the distinction between social and socio­mathematical norms is rather subtle, and does not provide a suf­ficiently clear basis for empirically grounded interpretations, because

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 319

most of the social norms discussed in the work of Cobb and his associ­ates are narrowly related to mathematical activities.

Second, Mottier Lopez has also rightly put forward that classroom norms and practices do not only result from the interactions among students and the teacher, but are also shaped by the broader context wherein the classroom is embedded (the school, the local community, the guidelines, and rules of the Department of Education) and that influences the standards of mathematics education, the subject matter to be taught, etc. (see also Greer et al., 2007). It should be mentioned, however, that Cobb and his co­workers themselves admit this restric­tion (e.g., Cobb et al., 2001).

Third, Cobb’s framework seems to be especially suitable for analyz­ing mathematics activities and practices in the context of design experi­ments that aim explicitly and systematically at installing new classroom norms. But according to the now available studies the framework looks less applicable for mapping and understanding in the context of ascer­taining studies how an already established classroom culture is con­tinuously reconfirmed, or how this culture and its possible subcultures emerge and develop in typical classrooms. In fact, Cobb et al. (2001) seem to be also aware of this limitation when acknowledging that in tra­ditional classrooms students are usually not expected to make explicit their thinking about and interpretations of the classroom norms and practices.

Fourth, and most importantly, the studies undertaken in the context of Cobb’s framework – including the investigation by Depaepe et al. (2007) discussed above – have so far not succeeded in linking and integrating in a systematic way their rich data and insights about the social processes of sense­making, communication, and negotiation in the mathematics classroom, on the one hand, with the development of the mathematics­related beliefs and cognitions of the individual children participating in these instructional settings, on the other hand. In this respect it has to be remarked that, whereas Cobb and associates have criticized the previ­ously dominant cognitive paradigm for overemphasizing the individual psychological perspective in analyzing classroom practices, in their own work they have reversely paid until now very much attention to the social perspective at the expense of the psychological approach (see also Vlas­sis, 2004). For instance, Yackel and Cobb (1996) focused in their study on a post hoc analysis of the classroom situation in order to identify particular social and socio­mathematical norms, whereas the individual perspective was kept in the background.

In view of better documenting the reflexive relationships between the classroom culture and students’ mathematics­related beliefs, there

Mathematics­related beliefs and the classroom culture320

is from a methodological perspective a need to construct instruments for assessing and measuring more thoroughly critical incidents of both the social as well as the individual dimensions of the classroom activi­ties and their interactions. For instance, in the study by Depaepe et al. (2007) discussed above the analysis of the classroom norms was only based on low­inference coding, which looks exclusively at concrete and externally observable activities and events during a lesson. As a con­sequence only the norms that were explicitly verbally stated by the teacher could be captured. Therefore, in order to also grasp the implic­itly negotiated norms that emerge and develop during teacher–student interaction, we attempt in an ongoing study to go beyond the low­inference approach by complementing it with high­inference coding, which aims at disclosing the deep structure of learning and teaching processes. According to Pauli et al. (2007), high­inference coding aims at identifying qualities of student–teacher interactions and classroom management based on mapping broader patterns of regularities during sequences of activities in a lesson against certain criteria, such as cog­nitive activation, motivational support, student orientation of teaching, etc. Moreover, in view of a more thorough assessment of the individ­ual aspects of sense­making and valuation of the classroom processes, especially the subjects’ mathematics­related ( epistemological) beliefs, we also use in this current study, besides questionnaires, in­depth interviews with the teachers and the students of the two participating classes.

A second methodological problem relates specifically to the interven­tion studies reviewed above. As already mentioned, due to the com­plexity of the interventions implemented in those investigations, it is impossible to establish the relative importance of the components of these learning environments in producing the positive impact on stu­dents’ beliefs and performance. Thus, this raises the methodological problem of confounding of variables. Here we touch on a major topic of debate in the recent literature, namely the potential of design experi­ments for educational research (e.g., Sandoval and Bell, 2004).

Educational implications

We end with some educational implications. As argued in the begin­ning of this chapter, the acquisition of a mathematical disposition is the ultimate goal of mathematics education for children from preschool until grade twelve. A major claim supported by the theoretical and empirical work discussed in the present chapter is that availing beliefs about mathematics and mathematics learning and problem­solving

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 321

constitute an important part of such a disposition and should thus be systematically pursued in students in an integrated way with the other components, namely knowledge and skills. We suggest that a success­ful pursuit of valuable student beliefs consists of a balanced use of two instructional roads, although we realize that the evidence for this rec­ommendation is quite scarce.

The first road consists of explicitly addressing during lessons the mathematics­related classroom norms that correspond to these val­ued beliefs, so as to make students aware of their often unavailing and incorrect beliefs as an essential step toward modifying and transform­ing them in a positive direction. This may include the explicit statement of such a norm by the teacher (e.g., as in the three aforementioned intervention studies, where the teacher sometimes explicitly stated and explained a norm, such as “The class as a whole will decide which of the generated solutions is the optimal one after an evaluation of the pros and cons of the different alternatives; this is not something that I will do alone” in the study of Verschaffel et al., 1999). Another way of explic­itly addressing mathematics­related classroom norms is by organizing discussions about these norms by groups of children or with the whole class, for instance starting from “debating cards” (English, 1998) that contain explicit beliefs about mathematics and mathematics learning and teaching (e.g., “You learn more from working on one hard problem than from working on ten easy problems” or “You can learn just as much from generating your own problems as you can from working the problems that your teacher gives you”).

Besides this explicit road to pursue valuable beliefs, which will typi­cally be chosen at the beginning of a new teaching/learning process or at times when the regular norms and practices are violated, there is another, implicit road, wherein students interiorize the valued norms in a more subtle and gradual way just by being immersed in the class­room culture and by participating in the regular classroom practices. For instance, the norm that solving a mathematical problem takes time cannot only be established by the teacher by explicitly stating and/or debating it, but also by actually giving students enough time for math­ematical problem­solving in regular class assignments and tests. Like­wise, the belief that there is no gap between mathematics and reality cannot only develop through explicit statements or debates of this issue, but also through regularly confronting students with genuine real­world problems that are present in newspapers, advertisements, etc., and allowing them to make realistic considerations during the mathe­matical modelling and solving process. And, finally, the establishment of the norm about how authority is distributed in the mathematics

Mathematics­related beliefs and the classroom culture322

classroom can be done explicitly (as in the above­mentioned exam­ple from Verschaffel et al.’s (1999) intervention study), but, again, also implicitly through effectively giving students more voice and agency in the selection of the problems to be solved, in the way in which the problem­solving activities are organized, and in the assessment criteria and procedures.

Curriculum designers and textbooks writers should also take all this into account and provide explicit guidelines for the development of a classroom culture that can facilitate the acquisition of positive beliefs in students. However, although this is necessary, it is, as shown in the study by Depaepe et al. (2007), by no means sufficient for guarantee­ing that in their teaching practices the teachers will spontaneously implement and establish in a sustainable way an appropriate classroom culture in which new norms are negotiated and developed in view of pursuing availing mathematics­related beliefs in learners. Teacher edu­cation and staff development will have to prepare and support teachers intensively in this respect, and this will often involve changing teach­ers’ own beliefs about mathematics teaching. As for their students, this change will also have to occur partly along an explicit and partly along an implicit road.

R EF ER ENCES

Baroody, A. J. and Dowker, A. (Eds.) (2003). The development of arithmetic concepts and skills: Constructing adaptive expertise. Mahwah, NJ: Lawrence Erlbaum Associates.

Belenky, M. F., Clinchy, B. M., Goldberger, N. R., and Tarule, J. M. (1986). Women’s ways of knowing: The development of self, voice, and mind. New York: Basic Books.

Bendixen, L. D. and Rule, D. C. (2004). An integrative approach to personal epistemology: A guiding model. Educational Psychologist, 39, 69–80.

Boaler, J. (1997). Experiencing school mathematics: Teaching styles, sex, and setting. Buckingham, UK: Open University Press.

(1998). Open and closed mathematics: Students experiences and under­standings. Journal for Research in Mathematics Education, 29, 41–62.

(1999). Participation, knowledge and beliefs: A community perspective on mathematics learning. Educational Studies in Mathematics, 40, 259–81.

(2000). Introduction: Intricacies of knowledge, practice, and theory. In J. Boaler (Ed.), Multiple perspectives on mathematics teaching and learning (1–17). Westport, CT: Ablex Publishing.

Boaler, J. and Greeno, J. (2000). Identity, agency, and knowing in mathematics worlds. In J. Boaler (Ed.), Multiple perspectives on mathematics teaching and learning (171–200). Westport, CT: Ablex.

Buehl, M. M. and Alexander, P. A. (2001). Beliefs about academic knowledge. Educational Psychology Review, 13, 385–418.

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 323

Buehl, M. M., Alexander, P. A., and Murphy, P. K. (2002). Beliefs about schooled knowledge: Domain specific or domain general? Contemporary Educational Psychology, 27, 415–49.

Carpenter, T. P., Lindquist, M. M., Matthews, W., and Silver, E. A. (1983). Results of the third NAEP mathematics assessment: Secondary school. Mathematics Teacher, 76, 652–9.

Cobb, P. (2000). Conducting teaching experiments in collaboration with teachers. In A. E. Kelly and R. A. Lesh (Eds.), Handbook of research design in mathematics and science education (307–33). Mahwah, NJ: Lawrence Erl­baum Associates.

Cobb, P., Hodge, L. L., Gresalfi, M. (2006). Students’ developing identities in mathematics classrooms. Paper presented at the Annual Meeting of the American Educational Research Association, San Francisco, CA (April).

Cobb, P., Stephan, M., McClain, K., and Gravemeijer, K. (2001). Participat­ing in classroom mathematical practices. Journal of the Learning Sciences, 10, 113–63.

Cobb, P. and Yackel, E. (1998). A constructivist perspective on the culture of the mathematics classroom. In F. Seeger et al. (Eds.), The culture of the mathematics classroom (158–90). Cambridge: Cambridge University Press.

Collins, A., Brown, J. S., and Newman, S. E. (1989). Cognitive apprentice­ship: Teaching the crafts of reading, writing, and mathematics. In L. Res­nick (Ed.), Knowing, learning, and instruction: Essays in honour of Robert Glaser (453–94). Hillsdale, NJ: Lawrence Erlbaum Associates.

De Bock, D., Van Dooren, W., Janssens, D., and Verschaffel, L. (2002). Improper use of linear reasoning: An in­depth study of the nature and the irresistibility of secondary school students’ errors. Educational Studies in Mathematics, 50, 311–34.

De Corte, E., Op ’t Eynde, P., and Verschaffel, L. (2002). Knowing what to believe: The relevance of mathematical beliefs for mathematics education. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychol-ogy of beliefs about knowledge and knowing (297–320). Mahwah, NJ: Law­rence Erlbaum Associates.

De Corte, E. and Verschaffel, L. (2006). Mathematical thinking and learning. In W. Damon et al. (Eds.), Handbook of child psychology, Vol. 4: Child psy-chology and practice (6th edn., 103–52). John Wiley and Sons.

De Corte, E., Verschaffel, L., Lasure, S., Borghart, I., and Yoshida, H. (1999). Real­world knowledge and mathematical problem solving in upper pri­mary school children. In J. Bliss et al. (Eds.), Learning sites: Social and technological contexts for learning (61–79). Oxford: Elsevier Science.

Depaepe, F., De Corte, E., and Verschaffel, L. (2007). Unraveling the culture of the mathematics classroom: A videobased study in sixth grade. Interna-tional Journal of Educational Research, 46, 266–79.

English, L. (1998). Problem posing in middle-school classrooms. Paper presented at the Annual Conference of the American Educational Research Association, San Diego (April).

Glaser, B. G. and Strauss, A. L. (1967). The discovery of grounded theory: Strate-gies for qualitative research. Chicago: Aldine Publishing Company.

Mathematics­related beliefs and the classroom culture324

Good, T. L., Mulryan, C., and McCaslin, M. (1992). Grouping for instruc­tion in mathematics: A call for programmatic research on small­group processes. In D. A. Grouws (Ed.), Handbook on mathematics teaching and learning (165–96). New York: Macmillan.

Gove, P. B. (1993). Webster’s third new international dictionary. Springfield, MA: Merriam­Webster, Inc.

Greeno, J. G. (1991). Number sense as situated knowing in a conceptual domain. Journal for Research in Mathematics Education, 22, 170–218.

Greer, B. and Verschaffel, L. (Eds.) (1997). Modelling reality in mathematics classrooms [special issue]. Learning and Instruction, 7, 293–397.

Greer, B., Verschaffel, L., and De Corte, E. (2002). “The answer is really 4.5”: Beliefs about word problems. In G. C. Leder et al. (Eds.), Beliefs: A hidden variable in mathematics education? (271–92). Dordrecht, The Netherlands: Kluwer Academic Publishers.

Greer, B., Verschaffel, L., and Mukhopadhyay, S. (2007). Modelling for life: Mathematics and children’s experience. In W. Blum, et al. (Eds.), Modelling and applications in mathematics education (ICMI Study 14) (89–98). New York: Springer.

Hofer, B. K. (2000). Dimensionality and disciplinary differences in personal epistemology. Contemporary Educational Psychology, 25, 378–405.

(2001). Personal epistemology research: Implications for learning and teach­ing. Educational Psychology Review, 13, 353–83.

(2002). Personal epistemology as a psychological and educational con­struct: An introduction. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (3–14). Mahwah, NJ: Lawrence Erlbaum Associates.

(Ed.) (2004). Personal epistemology: Paradigmatic approaches to under­standing student’s beliefs about knowledge and knowing. Educational Psy-chologist, 39, 1–80 (special issue).

Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learn­ing. Review of Educational Research, 67, 88–140.

(Eds.) (2002). Personal epistemology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

Institut de Recherche sur l’Enseignement des Mathématiques (IREM) de Gre­noble (1980). Bulletin de l’ Association des Professeurs de Mathématique de l’Enseignement Public, no. 323, 235–43.

Kloosterman, P. (1996). Students’ beliefs about knowing and learning mathe­matics: Implications for motivation. In M. Carr (Ed.), Motivation in math-ematics (131–56). Cresskill, NJ: Hampton Press.

(2002). Beliefs about mathematics and mathematics learning in the second­ary school: Measurement and implications for motivation. In G. C. Leder et al. (Eds.), Beliefs: A hidden variable in mathematics education? (247–69). Dordrecht, The Netherlands: Kluwer Academic Publishers.

Köller, O. (2001). Mathematical world views and achievement in advanced mathematics in Germany: Findings from TIMSS population 3. Studies in Educational Evaluation, 27, 65–78.

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 325

Lampert, M. (1990). When the problem is not the question and the solution is not the answer: Mathematical knowing and teaching. American Educa-tional Research Journal, 27, 29–63.

Leder, G. C., Pehkonen, E., and Törner, G. (Eds.) (2002). Beliefs: A hidden variable in mathematics education? Dordrecht, The Netherlands: Kluwer Academic Publishers.

Mason, L. and Scrivani, L. (2004). Developing students’ mathematical beliefs: An intervention study. Learning and Instruction, 14, 153–76.

McClain, K. and Cobb, P. (2001). An analysis of development of sociomath­ematical norms in one first­grade classroom. Journal for Research in Math-ematics Education, 32, 236–66.

McLeod, D. B. (1992). Research on affect in mathematics education: A recon­ceptualization. In D. A. Grouws (Ed.), Handbook of research on mathematics teaching and learning (575–96). New York: Macmillan.

Mottier Lopez, L. (2005). Co-constitution de la micro culture de classe dans une perspective située: Étude d’activités de résolution de problèmes mathématiques en troisième année primaire [Co­construction of a classroom micro cul­ture in a situated perspective: study of the problem­solving activities in a third­grade classroom]. Unpublished doctoral dissertation, University of Geneva, Switzerland.

Mottier Lopez, L. and Allal, L. (2007). Sociomathematical norms and the regulation of problem solving in classroom microculutres. International Journal of Educational Research, 46, 252–65.

Muis, K. R. (2004). Personal epistemology and mathematics: A critical review and synthesis of research. Review of Educational Research, 74, 317–77.

Muis, K. R., Bendixen, L. D., and Haerle, F. C. (2006). Domain­generality and domain­specificity in personal epistemological research: Philosophi­cal and empirical reflections in the development of a theoretical frame­work. Educational Psychology Review, 18, 3–54.

Nasir, N. S. (2002). Identity, goals, and learning: Mathematics in cultural practice. Mathematics Thinking and Learning, 4, 213–47.

National Council of Teachers of Mathematics (1989). Curriculum and evalu-ation standards for school mathematics. Reston, VA: National Council of Teachers of Mathematics.

(2000). Principles and standards for school mathematics. Reston, VA: National Council of Teachers of Mathematics.

National Research Council. (2001). Adding it up: Helping children learn mathematics. J. Kilpatrick, J. Swafford, and B. Findell (Eds.), Mathemat­ics Learning Study Committee, Center for Education, Division of Behav­ioral and Social Sciences and Education. Washington, DC: National Academy Press.

Nunes, T., Schliemann, A. D., and Carraher, D. W. (1993). Street mathematics and school mathematics. Cambridge: Cambridge University Press.

Op ’t Eynde, P., De Corte, E., and Verschaffel, L. (2002). Framing students’ mathematics­related beliefs: A quest for conceptual clarity and a compre­hensive categorization. In G. C. Leder et al. (Eds.), Beliefs: A hidden variable in mathematics education? (13–37). Dordrecht, The Netherlands: Kluwer Academic Publishers.

Mathematics­related beliefs and the classroom culture326

(2006). Epistemic dimensions of students’ mathematics­related beliefs. International Journal of Educational Research, 45, 57–70.

Pauli, C., Reusser, K., and Grob, U. (2007). Teaching for understanding and/or self­regulated learning? A videobased analysis of reform­oriented math­ematics instruction in Switzerland. International Journal of Educational Research, 46, 294–305.

Paulsen, M. B. and Wells, C. T. (1998). Domain differences in the episte­mological beliefs of college students. Research in Higher Education, 39, 365–84.

Pekhonen, E. (1995). Pupils’ view of mathematics: Initial report of an international comparison project. Research report 152. Helsinki: University of Helsinki, Department of Teacher Education.

Picker, S. H. and Berry, J. S. (2000). Investigating pupils’ images of mathema­ticians. Educational Studies in Mathematics, 43, 65–94.

Pintrich, P. R. (2002). Future challenges and directions for theory and research on personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Per-sonal epistemology: The psychology of beliefs about knowledge and knowing (389–414). Mahwah, NJ: Lawrence Erlbaum Associates.

Radatz, H. (1983). Untersuchungen zum Lösen eingekleideter Aufgaben. Zeitschrift für Mathematik-Didaktik, 4, 205–17.

Remillard, J. T. (2005). Examining key concepts in research on teachers’ use of mathematics curricula. Review of Educational Research, 75, 211–46.

Reusser, K. (1986). Problem solving beyond the logic of things: Contextual effects on understanding and solving word problems. Instructional Science, 17, 309–38.

Reusser, K. and Stebler, R. (1997). Every word problem has a solution: The suspension of reality and sense­making in the culture of school mathemat­ics. Learning and Instruction, 7, 309–28.

Sandoval, W. A. and Bell, P. (Eds.) (2004). Design­based research methods for studying learning in context [special issue]. Educational Psychologist, 39(4).

Schoenfeld, A. H. (1983). Beyond the purely cognitive: Belief systems, social cognition, and metacognitions as driving forces in intellectual perform­ance. Cognitive Science, 7, 329–63.

(1988). When good teaching leads to bad results: The disasters of “well­taught” mathematics courses. Educational Psychologist, 23, 145–66.

(1989). Exploration of students’ mathematical beliefs and behaviour. Journal for Research in Mathematics Education, 20, 338–55.

(1991). On mathematics as sense­making: An informal attack on the unfor­tunate divorce of formal and informal mathematics. In J. F. Voss et al. (Eds.), Informal reasoning and education (311–43). Hillsdale, NJ: Lawrence Erlbaum Associates.

(1992). Learning to think mathematically: Problem solving, metacogni­tion, and sense­making in mathematics. In D. A. Grouws (Ed.), Hand-book of research on mathematics teaching and learning (334–70). New York: Macmillan.

E. De Corte, P. Op ’t Eynde, F. Depaepe, and L. Verschaffel 327

Schommer, M. (1994). Synthesizing epistemological belief research: Tenta­tive understandings and provocative confusions. Educational Psychology Review, 6, 293–319.

Schommer­Aikins, M. (2002). An evolving theoretical framework for an epis­temological belief system. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (103–18). Mahwah, NJ: Lawrence Erlbaum Associates.

Seeger, F., Voigt, J., and Waschescio, U. (1998a). Introduction. In F. Seeger et al. (Eds.), The culture of the mathematics classroom (1–9). Cambridge: Cam­bridge University Press.

(Eds.) (1998b). The culture of the mathematics classroom. Cambridge: Cam­bridge University Press.

Stodolsky, S. S., Salk, S., and Glaessner, B. (1991). Student views about learn­ing math and social studies. American Educational Research Journal, 28, 89–116.

Underhill, R. (1988). Mathematics learners’ beliefs: A review. Focus on Learn-ing Problems in Mathematics, 10, 55–69.

Verschaffel, L., De Corte, E., and Lasure, S. (1994). Realistic considerations in mathematical modelling of school arithmetic word problems. Learning and Instruction, 4, 273–94.

Verschaffel, L., De Corte, E., Lasure, S., Van Vaerenbergh, G., Bogaerts, H., and Ratinckx, E. (1999). Learning to solve mathematical application problems: A design experiment with fifth graders. Mathematical Thinking and Learning, 1, 195–229.

Verschaffel, L., Greer, B., and De Corte, E. (2000). Making sense of word prob-lems. Lisse, The Netherlands: Swets and Zeitlinger.

Vlassis, J. (2004). Sens et symboles en mathématiques [Meaning and symbols in mathematics]. Thèse non publié. Liège, Belgium: Université de Liège, Faculté de Psychologie et des Sciences de l’Education.

Wenger, E. (1998). Communities of practice: Learning, meaning and identity. Cambridge: Cambridge University Press.

Yackel, E. and Cobb, P. (1996). Sociomathematical norms, argumentation, and autonomy in mathematics. Journal for Research in Mathematics Education, 27, 458–77.

328

11 Examining the influence of epistemic beliefs and goal orientations on the academic performance of adolescent students enrolled in high­poverty, high­minority schools

P. Karen Murphy The Pennsylvania State University

Michelle M. Buehl George Mason University

Jill A. Zeruth The Pennsylvania State University

Maeghan N. Edwards University of Oklahoma

Joyce F. Long The University of Notre Dame

Shinichi Monoi The Ohio State University

Introduction

Educational and psychological research over the last decade has consist­ently supported the presupposition that learning is a dynamic process in which knowledge and motivation work in concert to influence achieve­ment (Alexander, 1997; Pintrich et al., 1993; Tanaka and Yamauchi, 2001). Research by Schommer (e.g., 1990, 1993) and others (e.g., Hofer, 2000; Kardash and Scholes, 1996; King and Kitchener, 1994) has also shown that students’ beliefs about knowledge and the process of know­ing (i.e., epistemology) play powerful roles in their learning and develop­ment. In addition, it is now understood that constructs like knowledge, motivation, and epistemology are multidimensional and make unique contributions to the learning process (e.g., Alexander, 1997; Buehl et al., 2002; Middleton and Midgley, 1997; Schommer, 1990). For example, research investigating goal orientations in ninth­grade students has shown that avoidance­oriented goals are correlated with higher test anxiety, while performance­oriented goals are related to lower self­ efficacy (Niemivirta, 2002). Additional studies suggest that these rela­tions may vary according to developmental levels. Whereas learning or mastery goals relate to both “higher levels of content knowledge and better grades” among middle school students (Gehlbach, 2006, p. 366),

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 329

the process becomes more differentiated in college. Specifically, per­formance goals consistently correlate with grades while learning goals appear to be more closely aligned with student interest (Harackiewicz et al., 2005).

Although much is known about the role constructs like epistemic beliefs and achievement goals play in students’ learning, very little research has explored how these factors work in concert to influence educational performance. This is especially true for students from minority groups within high­poverty districts. Despite the fact that motivational factors “are at the heart of contemporary concerns about the status of African­Americans in general and their academic achieve­ments in particular” (Graham, 1994, p. 55), our understanding of how knowledge and motivation partner to support learning in urban class­rooms is very limited.

Sociologists addressing the underperformance of African­ American students relative to their Caucasian counterparts (Steinberg et al., 1992) have formulated several explanations. For example, students of color may choose to willfully oppose academic actions associated with being white (Fordham and Ogbu, 1986) or struggle with negative stereotypes that disparage their intellectual abilities (Steele, 1997). In addition, individuals in school districts where the majority of its students are impoverished face other challenges because of their low socio­economic status. These include transiency (Beck et al., 1997), academic and dis­cipline disparities (Lipman, 2004), limited access to knowledge (Oakes, 2005), lack of resources (Kozol, 2005), and low expectations (Finn, 1999). However, when these factors occur simultaneously with stressful academic transitions (Reyes et al., 2000; Seidman et al., 1994), stu­dent outcomes are even more likely to be negatively affected (Long et al., 2007). Thus, any model attempting to account for the academic achievement of African­American students must attend to multiple influences and factors (Graham, 1994).

In this chapter, we report on a portion of a larger data set inves­tigating the difficulties of the transition from middle school to high school for adolescents attending high­poverty, high­minority schools. We attempt to broaden current understandings about the roles of epis­temic beliefs and academic achievement goals in the learning process in two ways. First, we examine the dimensionality of the goal orienta­tions for eighth­ and ninth­grade students enrolled in a high­poverty, high­minority school. Second, we explore the influence of epistemic beliefs and achievement goals on the academic achievement of these adolescents. To our knowledge, no other investigation has examined the beliefs and goals of students enrolled in these types of classrooms.

Epistemic beliefs and high­poverty/high­minority schools330

As such, it is unknown whether well­researched, psychometrically sound measures (e.g., Middleton and Midgley, 1997) will retain those properties when employed with students enrolled in diverse settings. Similarly, very little is known about the influence of epistemic beliefs and goals for learning on students’ classroom performance.

Epistemic beliefs

Research within the burgeoning epistemic belief literature has focused on exploring: (1) the nature and dimensionality of students’ epistemic beliefs (e.g., Perry, 1970; Schommer, 1990); (2) the domain­specificity of epistemic beliefs (e.g., Buehl et al., 2002; Hofer, 2000); (3) the develop­ment of epistemic beliefs over time (e.g., Baxter Magolda, 1992; Perry, 1970); and (4) the relations between epistemic beliefs and learning out­comes (e.g., Hofer, 2000; Qian and Alverman, 1995). In the present investigation, we are most interested in the nature of epistemic beliefs, epistemic belief development over time, and the relations between epis­temic beliefs and academic performance (e.g., Hofer, 2000; Qian and Alverman, 1995).

Nature of epistemic beliefs. There is a long history of research exam­ining the nature of beliefs about knowing and learning beginning with Perry (1970). Based on his initial work with college students, Perry (1970) concluded that students’ beliefs about knowledge fall along a continuum from dualism to relativism. Others have adopted a similar perspective and proposed categories or levels of beliefs along a single continuum (e.g., Baxter Magolda, 1992; King and Kitchener, 1994). In contrast, Schommer (1990) suggested students’ epistemic beliefs may be multidimensional, consisting of several independent belief factors. Drawing on previous belief research (e.g., Dweck and Leggett, 1988; Perry, 1970; Schoenfeld, 1983; Schommer, 1990) she proposed five dif­ferent epistemological dimensions related to the structure, certainty, and source of knowledge, as well as speed and control in the acquisition of knowledge.

Schommer and her colleagues empirically validated a four­factor structure of epistemic beliefs with college and high school students (e.g., Schommer, 1990, 1993; Schommer et al., 1997; Schommer et al., 1992). Those factors were initially referred to as innate ability, quick learning, simple knowledge, and certain knowledge, and were later renamed ability to learn, speed of learning, structure of knowledge, and stability of knowl-edge (Schommer­Aikins et al., 2000). Essentially, the factors pertain to students’ beliefs about whether (1) the ability to acquire knowledge is innate or acquired (i.e., ability to learn); (2) learning occurs quickly or

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 331

gradually over time (i.e., speed of learning); (3) knowledge is isolated and simplistic or complex and well­integrated (i.e., structure of knowledge); and (4) knowledge is certain and unchanging or tentative and evolving (i.e., stability of knowledge).

Although Schommer’s measure and conceptualization of beliefs has been used extensively, her work has also been criticized. For instance, Hofer and Pintrich (1997) argued that beliefs about ability and the speed of learning are more related to beliefs about learning and intelli­gence than beliefs about the nature of knowledge. Such criticisms have lead to modifications of Schommer’s measure as well as the develop­ment of new measures and alternative conceptualizations of epistemic beliefs (e.g., Buehl et al., 2002; Hofer, 2000; Jehng et al., 1993; Qian and Alverman, 1995; Schraw et al., 2002). However, much of the afore­mentioned research has focused on the beliefs of high school and col­lege students.

Investigations of younger populations reveal that middle school stu­dents’ beliefs may be less differentiated. For instance, Schommer­Aikins and her collaborators administered a modified version of Schom­mer’s instrument to 1,296 seventh­ and eighth­grade students (i.e., Schommer­Aikins et al., 2005; Schommer­Aikins et al., 2000). Using confirmatory factor analyses to determine whether the factor struc­ture identified for college and high school students could be used with the middle school population, Schommer­Aikins et al. (2000) found that the initial four­factor model did not fit the data well. An alter­native three­factor model was developed with belief factors related to ability to learn, speed of learning, and stability of knowledge (Schommer­Aikins et al., 2000). A later exploratory factor analysis of the data identified a four­factor solution (Schommer­Aikins et al., 2005). How­ever, in both Schommer­Aikins et al. (2000) and Schommer­Aikins et al. (2005), the reliability coefficients for factors related to beliefs about knowledge were unacceptable (i.e., < .55), whereas the reliabil­ity coefficients were acceptable for data associated with beliefs about learning. Schommer­Aikins and colleagues suggested that middle school students’ beliefs are simpler and less differentiated than older students and that the development of learning beliefs precedes the development of knowledge beliefs (Schommer­Aikins et al., 2000; Schommer­Aikins, 2004).

Epistemic beliefs development over time. The development of epistemic beliefs has also been investigated from cross­sectional and longitudi­nal perspectives. As was the case with college students (e.g., Baxter Magolda, 1992; Perry, 1970), Schommer­Aikins’ studies suggested that as adolescents grow older, their epistemic beliefs become more

Epistemic beliefs and high­poverty/high­minority schools332

sophisticated (e.g., they believed less in quick learning). For example, high school seniors believed less in simple learning, certain knowledge, and fixed ability than younger students (Schommer, 1993; Schommer et al., 1997). Kuhn and colleagues (e.g., Kuhn et al., 2000) have found similar developmental trends in individuals’ epistemic judgments from fifth grade through adulthood. In essence, individuals’ epistemic judg­ments tend to be more absolute earlier in life and then transition toward a more evaluative position in adulthood. However, Kuhn and Wein­stock (2002) described the differences across their participant groups as modest.

Epistemic beliefs and academic performance. Due to our interest in the linkages between students’ epistemic beliefs and achievement, we reviewed relevant literature, paying particular attention to moderating student characteristics (e.g., ethnicity, age, or socio­economic status). Conley, Pintrich, Vekiri, and Harrison (2004) is the only study we iden­tified that examined epistemic beliefs, academic performance, and stu­dent socio­economic status or ethnicity. Specifically, in their sample (46 percent Latino, 27 percent Anglo, and 27 percent African­American; 67 percent free or reduced lunch) of fifth­grade students, Conley et al. (2004) found that beliefs about science knowledge were related to stu­dents’ academic achievement. Additionally, socio­economic status was negatively related to students’ beliefs, but ethnicity was not related to students’ beliefs. However, Conley et al. did not investigate the extent to which the relation between beliefs and achievement was moderated by socio­ecomonic status or ethnicity.

Few other studies have examined adolescents’ epistemic beliefs in relation to their academic achievement (e.g., Schommer and Dunnell, 1994), and these studies focused almost exclusively on students of European­American, middle­class heritage. For example, Schommer (1990, 1993) found that at the college and high school levels, students’ beliefs related to innate ability, quick learning, simple knowledge, and certain knowledge significantly predicted students’ grade point aver­age even after controlling for students’ verbal intelligence quotient (IQ). For seventh­ and eighth­grade students, beliefs about the abil-ity to learn and speed of learning were related to performance such that students who believed the ability to learn was innate or that learning should occur quickly tended to have lower grade point averages (GPAs) ( Schommer­Aikins et al., 2000). However, the participants in these investigations were primarily European­American adolescents whose socio­economic status was middle class, varied, or unknown. Thus, the generalizability of the aforementioned outcomes to students who are enrolled in high­poverty, high­minority schools is unknown.

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 333

Achievement goals

Within the motivation literature, students’ achievement goal orienta­tions have received considerable attention (Pintrich, 2000; Urdan and Midgley, 2003; Wolters, 2004). Researchers are interested in how stu­dents’ reasons or purposes for engaging in learning­oriented activities are related to their choice of, and persistence at, academic tasks as well as their performance (e.g., Ames and Archer, 1988; Dweck and Leggett, 1988). In their initial discussion of the nature and role of achievement goals, Dweck and Leggett (1988) proposed that students’ goals are influenced by their belief systems. Because students’ epistemic beliefs represent an aspect of their belief systems that are particularly relevant in academic settings, we were interested in how students’ goal orienta­tions are related to their beliefs about knowledge and their academic performance. In addition, in the present study, we are particularly interested in the extent to which existing understandings regarding the dimensionality of students’ goals hold for students enrolled in high­poverty, high­minority schools.

Nature of achievement goals. Several conceptually and empirically distinct types of achievement goal orientations have been identified. Initially, researchers distinguished between two main types of goal orien­tations: learning goals and performance goals (Ames, 1992; Blumenfeld, 1992; Dweck and Leggett, 1988). Learning goals pertain to an individual’s desire to develop competence and increase knowledge and understanding through effortful learning (Ames and Archer, 1988). A performance goal orientation reflects a desire to gain favorable judgments and avoid nega­tive judgments of one’s competence, particularly if success is achieved through a minimum exertion of effort (Dweck, 1986).

More recently, the distinctions have been made between two types of performance goals (i.e., performance­approach and performance­avoidance goals, Elliot and Harackiewicz, 1996). Performance­approach goals pertain to one’s desire to complete tasks that would allow the indi­vidual to appear competent and receive praise. Performance­avoidance goals relate to one’s desire to complete tasks that allow the individual to avoid appearing incompetent or to avoid receiving negative feedback. Distinct performance­approach and performance­avoidance goals have been identified with various ages (e.g., college: Elliot and Harackiewicz, 1996; middle school: Middleton and Midgley, 1997) and in diverse countries (e.g., Norway: Skaalvik, 1997; United States: Pajares et al., 2000). In particular, Middleton and Midgley (1997) empirically vali­dated two performance goals using confirmatory factor analytic tech­niques on data from 703 sixth­grade students. However, the correlation

Epistemic beliefs and high­poverty/high­minority schools334

between these two factors was significant (r = .56), raising questions of overlap between the two constructs. Similar results have been found by other researchers (e.g., Midgley and Urdan, 2001; Pajares et al., 2000) in varied content areas (e.g., writing and science). In all of these studies, high correlations have been found between the performance­approach and performance­avoidance factors, thus raising questions about the similarities of these two factors.

In addition to learning and performance goal orientations, a third, lesser investigated category has also been discussed: work­avoidant goals. For this orientation, students’ main concern is to get work done with a minimum amount of effort (Meece et al., 1988). In a study by Meece and Holt (1993) with fifth­ and sixth­grade students, work­avoidant goals were used as a measure of avoidance motivation and were nega­tively correlated with mastery goals (rs = ­.14 to ­.48). Work­avoidant goals were not included in the aforementioned studies, nor were per­formance goals included in the study by Meece and Holt (1993). Thus, the extent to which performance­avoidance and work­avoidant goals are related remains an empirical question. The performance­avoidance construct in these more recent studies (e.g., Pajares et al., 2000) may be explaining variance that was once explained by work­avoidant goal orientations (e.g., Meece et al., 1988).

Influence of achievement goal orientations on academic performance. As was the case with epistemic beliefs, we were particularly attuned to issues of student characteristics and their moderating effects on the relationship between achievement goals and academic performance. Much of the research we reviewed on goal orientations provided infor­mation on students’ ethnicity and even focused on this construct as a possible moderating variable (e.g., Pajares et al., 2000). Aside from ethnicity, however, those studies offered very little specific informa­tion regarding the participants. Based on the reviewed studies, it would appear that few, if any, researchers have explored the goal orientations of students enrolled in high­poverty, high­minority schools.

In reviewing literature on the relations between the goal orienta­tions and academic performance of young adolescents, we found sev­eral trends. First, there appears to be a weak, negative relation between achievement goals and academic performance. For example, in the Middleton and Midgley (1997) study, the correlations between sixth­grade students’ performance­approach goals, performance­avoidance goals, and academic performance were weak and negative (correlations ranging from –.02 to –.18). Pajares et al. (2000) reported similar results for sixth­, seventh­, and eighth­grade students of varied ethnicities. Specifically, only performance­avoidance goals significantly predicted

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 335

academic performance (r = –.27). Moreover, these researchers showed that African­American students reported modestly stronger per­formance­avoidance goals than did European­American students. In addition, students with relatively low prior knowledge in science tended to hold a performance­avoidance goal orientation regardless of ethnicity.

There is also some evidence that mastery goals can affect the achieve­ment of students from underrepresented populations during adoles­cence. Guttman (2006) found that African­American students who were more likely to endorse mastery goals in ninth­grade experienced more positive change in mathematics achievement from eighth­ to ninth­grade than those who were less likely to endorse mastery goals, after controlling for endorsement of mastery goals in the eighth­grade.

Relations between epistemic beliefs and achievement goals

Similar to other motivational variables, achievement goals are related to choice of activity, persistence, and strategy use (Pintrich, 2000). In addition, goal orientations are believed to relate to students’ epistemic beliefs. According to Dweck and Leggett (1988), students’ motivational goals are influenced by their belief systems. In their work, they have shown that students’ beliefs about intelligence affect their choice of learning or performance goals (Dweck et al., 1982). Students who were manipulated to believe that the probability of learning in particular experiments was fixed tended to adopt performance goals. In contrast, students who were led to believe that intelligence could be increased tended to adopt learning goals. In essence, these researchers found that beliefs affect students’ goal orientations.

We hold that students’ beliefs about knowledge and learning repre­sent another aspect of their belief system that may influence their goal orientations. There is also empirical support for this proposed relation. For example, Qian and Burrus (1996) found that students’ beliefs about speed of learning, certainty/simplicity of knowledge, and innateness of ability were significant predictors of a performance or learning goal orienta­tion. However, Qian and Burrus (1996) conceptualized goal orientation as existing along a single continuum, ranging from learning goal to per­formance goal, making it difficult to delineate specific relations among the belief factors and goal orientations.

Significant relations have also been shown between epistemic beliefs and intrinsic or extrinsic goals. For instance, Paulsen and Feldman (1999) found that college students’ beliefs about the simplicity of knowl-edge were positively related to extrinsic goals and negatively related to

Epistemic beliefs and high­poverty/high­minority schools336

intrinsic goals. Beliefs about the speed of knowledge acquisition and fixed ability were also negatively related to intrinsic goals. Further, in her study of epistemic beliefs and motivation, Hofer (1999) found a signifi­cant correlation between students’ beliefs about mathematics knowl­edge and their intrinsic goals. Specifically, the less students believed mathematics was an isolated activity, the more they tended to have an intrinsic goal orientation. Additional research is needed to investigate how students’ epistemic beliefs may relate to goal orientations and aca­demic performance.

Current investigation

Despite the depth of research already present in the literature on epistemic beliefs and achievement goals, very little is known about the structure of these constructs for young adolescents attending high­poverty, high­minority schools. In addition, there are virtually no studies, to date, that explore the predictive powers of these con­structs on academic achievement for such a population. As such, the overarching purpose of this investigation was to explore the epistemic beliefs and achievement goals of eighth­ and ninth­grade adolescents attending schools characterized by high poverty and a proportionally large minority population. To accomplish this purpose, we exam­ined: (1) the dimensionality of students’ goal orientations; and (2) the influence of epistemic beliefs and goal orientations on students’ aca­demic achievement. Based on the existing literature, we anticipated that distinct goal orientations would emerge related to learning goals, performance­approach goals, performance­avoidance goals, and work­avoidant goals. Further, we hypothesized that students’ epistemic beliefs would predict goal orientations which would, in turn, predict students’ academic achievement.

Method

Participants

Participants for this study included 450 adolescent students in eighth­ and ninth­grades (218 girls and 232 boys) in a large, urban, inner­city school district in the Midwestern United States. Eighth­grade respond­ents attended three middle schools serving as feeder schools for the high school from which the ninth­grade sample was collected. Conse­quently, 57 of the students (i.e., 14 percent of the total sample: 34 girls and 23 boys) participated when they were in both the eighth and ninth

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 337

grades. However, we did not view this as problematic since all analyses were conducted separately on the eighth­ and ninth­grade data.1

For the eighth­grade sample, 255 students (132 girls and 123 boys) participated, ranging in age from thirteen to sixteen years with a mean age of fourteen years. Of the students in the sample, 90 percent were members of a minority group, with 87 percent African­American, 10 percent Caucasian, and 3 percent from other minority groups (e.g., Asian, Native­American, and Multi­Ethnic). The extent to which stu­dents were enrolled in free or reduced lunch programs was used as a proxy for socio­economic status. This data was available for 70 percent of the participants in the eighth­grade sample, with 70 percent of this group receiving free or reduced lunch. That is, at least 125 eighth­grade students (i.e., 49 percent of the eighth­grade sample) were enrolled in the free and reduced lunch program. The number of students in this program was most likely higher but could not be determined due to lack of available data at the school level.

In addition, 195 ninth­grade students (86 girls and 109 boys) partici­pated, ranging in age from fourteen to nineteen years with a mean age of fifteen years. Of these students, 80 percent identified themselves as being part of a minority group, with 64 percent African­American, 20 percent Caucasian, 10 percent Multi­Ethnic, 3 percent Asian, 2 per­cent Hispanic, and 1 percent Native­American. The extent to which students were enrolled in free or reduced lunch programs was again used as a proxy of socio­economic status. However, this data was only available for a portion of our ninth­grade sample (i.e., 75 percent of the ninth­grade participants). Based on the available data at least 88 ninth­grade students (i.e., 45 percent of the ninth­grade sample) were enrolled in the free and reduced lunch program. Again, we believe the actual number may be higher.

Regardless of the exact numbers of students enrolled in the free and reduced lunch program, the available data indicated that these students are enrolled in schools with a high occurrence of poverty (i.e., almost half of both the eighth­ and ninth­grade samples) and comprised of students representing diverse minority populations (i.e., 90 percent of the eighth­grade sample and 80 percent of the ninth­grade sample). These characteristics differentiate these samples from those studied in

1 Although these data are part of a larger data set focused on the transition from eighth­ to ninth­grade, we analyzed the data cross­sectionally in the current manuscript rather than longitudinally. This decision was based on the fact that dimensionality of students’ responses to the epistemic belief measure changed over time (Murphy et al., 2007). Further, only 14 percent of the sample appears in both the eighth­ and ninth­grade.

Epistemic beliefs and high­poverty/high­minority schools338

previous investigations. Indeed, our review revealed that most investi­gations focused on Caucasian, middle­class populations.

Measures

Epistemic beliefs. In this study, students’ epistemic beliefs were assessed using an adapted version of the domain-specific belief questionnaire (DSBQ). The DSBQ was originally developed and validated with college­age students by Buehl et al. (2002). For the Buehl et al. col­lege sample, twenty­two of the original fifty items on the DSBQ loaded highly onto one of four factors: need for effort in mathematics (five items; α = .68; example item: “There is a relationship between the number of hours students study and how well they do in mathematics”); integra-tion of information and problem-solving in mathematics (six items; α = .70; example item: “There are links between mathematics and other disci­plines”); need for effort in history (five items; α = .61; example item: “Stu­dents who are good at history have to work hard”); and integration of information and problem solving in history (six items; α = .75; example item: “It is important for students to integrate new ideas in history with what they already know”). Participants indicated the extent to which they agreed with an item on a ten­point Likert­type scale anchored with strongly disagree and strongly agree.

As discussed in prior research (Murphy et al., 2007), confirmatory factor analyses revealed that the factor structure identified in Buehl et al. (2002) was not appropriate for these adolescents’ beliefs about knowledge. Subsequent exploratory factor analyses revealed different factor structures for eighth­ and ninth­graders. In the eighth­grade sample, forty­three items loaded highly on one of two factors: beliefs about learning (thirty­five items; pattern coefficients ranging from .41 to .83) and reverse-coded items (eight items; .43 to .84). Murphy et al. (2007) surmised that these younger students were unable to process the negatively worded items beyond the negative surface structure. The Cronbach alpha reliability for the data associated with these two factors were .94 and .75, respectively.

Additional belief dimensions emerged for scores from the ninth­grade sample (Murphy et al., 2007). Specifically, forty­two items loaded highly on one of three factors: beliefs about learning (23 items; pattern coefficients ranging from .42 to .82); reverse-coded items (11 items; .42 to .80); and beliefs about work (eight items; .57 to .81). Moreover, Murphy et al. (2007) found that the ninth­graders, like their eighth­grade coun­terparts, had difficulty deeply processing the meaning of the reverse­coded items. Reported reliabilities ranged from .76 to .93. Further, in

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 339

contrast to Buehl et al. (2002), domain­specific factors did not emerge in the eighth­ nor ninth­grade data (Murphy et al., 2007).

Although the factor structure in Murphy et al. (2007) differed from initial expectations, data for the emergent factors were reli­able and addressed different beliefs about learning and knowledge. Thus, we felt it was a reliable and valid measure of students’ beliefs. The notable exception was the reverse-coded items factor. Results from Murphy et al. (2007) suggested that adolescent students have diffi­culty with negatively worded items. As such, we did not include these items in the analyses for the present investigation. The epistemic beliefs measure in the current study contained thirty­five items for the eighth­grade sample and thirty­one items for the ninth­grade sample on a ten­point Likert­type scale ranging from strongly dis-agree to strongly agree. Table 11.1 presents a list of the epistemic belief items used for this study along with the factors on which they loaded ( Murphy et al., 2007).

Achievement goal orientations. Students’ achievement goal orientations were measured in this study using twenty­four items based on previ­ous measures validated in the literature. Questions evaluating students’ learning, performance­approach, and performance­avoidance goals were assessed with three domain­general goal orientation scales (six items each) adapted from Midgley et al. (1998). These three achieve­ment goal orientations were first examined with math­specific scales (Middleton and Midgley, 1997), which were developed from the pat­terns of adaptive learning survey (PALS; Midgley et al., 1996). The psychometric properties of these scales were reported by Midgley et al. (1998) based on data from middle school students. Confirmatory fac­tor analyses indicated that the three­factor goal orientation model pro­vided the best fit for the data of boys and girls of African­American and European­American descent. Pattern coefficients ranged from .42 to .81 for the three goal factors. Reported Cronbach alpha coefficients for learning, performance­approach, and performance­avoidance goals were .83, .86, and .74, respectively. The original scales for the three goals were assessed using five­point Likert scales. In the present study, the scale was adapted to a ten­point Likert­type scale to make scales of measurement consistent among all variables. Also, the extended scale allows for the possibility of greater variability in the data and higher levels of reliability. The fourth achievement goal orientation, work­avoidant, was assessed using a six­item scale, adapted from previ­ous work by Meece and colleagues (Meece et al., 1988). Prior reliabilities with younger adolescents were not available. As with questions for the three other goal orientations, a ten­point Likert­type scale was used.

Epistemic beliefs and high­poverty/high­minority schools340

Table 11.1. Items assessing each of the factors for eighth- and ninth-grade participants

Eighth grade Ninth grade

Items

Beliefs about learning

Beliefs about learning

Beliefs about work

8. Someone can teach a student how to learn math material.

X X

9. Information learned in history is useful outside of school.

X X

14. Even if it takes a long time to learn a history concept, it is best to keep trying.

X X

16. Reviewing the material discussed in class would help a student learn math.

X X

18. Someone can teach a student how to learn history material.

X X

27. A course in history study skills would be valuable.

X X

29. If a student is not naturally gifted in history, they can still learn the content well.

X X

30. Mathematics relates to day to day life.

X X

31. A course in math study skills would be valuable.

X X

35. Even if it takes a long time to learn a math concept, it is best to keep trying.

X X

36. A math problem can be approached in several different ways.

X X

38. There are methods for learning mathematics.

X X

40. Reviewing the material discussed in class would help a student learn history.

X X

43. If a student is not naturally gifted in mathematics, they can still learn the content well.

X

X

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 341

Eighth grade Ninth grade

Items

Beliefs about learning

Beliefs about learning

Beliefs about work

44. Information learned in mathematics is useful outside of school.

X X

45. It is important for students to integrate new ideas in math with what they already know.

X X

48. Exerting effort to try to understand a tough problem in history is a wise use of time.

X X

6. How successful students are in history is related to how hard they work.

X X

7. There are links between mathematics and other disciplines.

X X

12. How successful students are in mathematics is related to how hard they work.

X X

15. History relates to day to day life. X X

21. There are links between history and other disciplines.

X X

22. It is important for students to integrate new ideas in history with what they already know.

X X

23. Exerting effort to try to understand a tough problem in math is a wise use of time.

X X

25. Students who are good at math have to work hard.

X X

2. Strategies will help a student learn mathematics.

X

4. A history question can be approached in several different ways.

X

10. Strategies will help a student learn history.

X

17. Most people can do well in history. X

Table 11.1. (cont.)

Epistemic beliefs and high­poverty/high­minority schools342

Eighth grade Ninth grade

Items

Beliefs about learning

Beliefs about learning

Beliefs about work

19. There is a relationship between the number of hours students study and how well they do in mathematics.

X

24. There are methods for learning history.

X

33. There is a relationship between the number of hours students study and how well they do in history.

X

49. Most people can do well in mathematics.

X

1. Students who are good at history have to work hard.

X

20. Students need repeated exposure to the material to understand math concepts.

X

Table 11.1. (cont.)

Procedures

The questionnaires were administered in a classroom setting during the spring of two sequential years (i.e., eighth­grade and then ninth­grade the following year). Given the difficulty in assessing young stu­dents, various means were employed to enhance student cooperation and the consistency of data collection. For instance, students were allowed to take as much time as needed to complete the measure, and most students completed the measures within forty­five minutes. Also, scripts were used for the administration of all of the measures, and were read aloud by trained research assistants to provide consistency across the two years. Given the diversity of the student population, we felt it was important that those individuals collecting data mirrored the ethnic diversity of the students. Specifically, the data collection team was comprised of four graduate students of varying ethnicities and both genders (i.e., one Asian male, one African­American male, two Caucasian females). In addition, the lead research assistant was in the students’ classrooms regularly throughout the school year and the students were familiar with her. To enhance compliance during data

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 343

collection all research assistants and the classroom teacher were pre­sent during data collection in regular classrooms. In addition, data for individuals missing more than five responses on a single measure were discarded from further analysis. Participating students’ grades for the spring semester (i.e., reading, history, math, and science) and lunch status were obtained directly from students’ records after the school year ended.

The data collection procedures spanning two years resulted in fifty­seven students participating in the study in both the eighth­ and ninth­grades. To determine whether students participating in both years came from the same population as those students participating in only one year, Hotelling’s T 2 tests were performed. Data from these fifty­seven students was compared with data from a randomly selected fifty­seven students who participated only once for each of the eighth­ and ninth­grades. These analyses were performed separately.

To determine whether scores were the same across the two types of participants (i.e., longitudinal and non­longitudinal), composite scores were calculated for two factors (i.e., overall beliefs and overall motiva­tion). These composite scores were calculated by adding ratings for all items employed to measure students’ beliefs or goals. Hotelling’s T2 tests were then completed to determine whether the vectors of mean scores across both types of participants were significantly different. Compos­ite scores for beliefs and motivation were treated as dependent variables with the longitudinal group (i.e., longitudinal and non­longitudinal) as the independent variable.2

For the eighth­grade sample, the mean vectors for these two groups of participants were not statistically significantly different (T2 = .755, F2, 111 = .374, p = .689). This tells us that none of the means for each of the composite scores were significantly different between the two groups, giving evidence that the eighth­grade students in longitudinal and non­longitudinal groups came from the same population. Similar results were found for ninth­grade participants. Hotelling’s T2 showed that there were no statistically significant differences in the vectors of mean scores for the two groups (T2 = .361, F2, 111 = .179, p = .836). We can conclude from this that the ninth­grade students in both groups were drawn from the same population. Thus, data for the longitudinal and non­longitudinal cases were collapsed and all data from each grade were analyzed together for the remainder of the analyses.

2 Hotelling’s T2 tests were also conducted after all other analyses with composite scores for each of the factors. These tests showed similar results, and are available upon request.

Epistemic beliefs and high­poverty/high­minority schools344

Results and discussion

In this investigation, we explored the dimensionality of goal orienta­tions for adolescents enrolled in high­poverty, high­minority schools as well as the influence of both their epistemic beliefs and goal orientations on academic achievement. Given the variations in the dimensionality of epistemic beliefs for the eighth­ and ninth­grade students identified in our assessment of the psychometric properties of the epistemic beliefs measure (Murphy et al., 2007), we chose to analyze the data from each sample separately in this investigation. For each grade level, we first conducted factor analyses to determine the dimensionality of students’ scores from the goal orientation measure. We then conducted path analyses to explore the relations among students’ epistemic beliefs, goal orientations, and academic achievement.

Dimensionality of students’ goal orientations

Eighth-grade goal structure. To determine the best model for the eighth­grade goal orientations data, confirmatory factor analyses were first conducted, because the instrument we employed was conceptually defined to measure four different goal orientations (i.e., learning goals, performance approach goals, performance avoidance goals, and work avoidance goals). However, the fit of this four­factor model to the eighth­grade data was marginal according to Hu and Bentler’s (1998) fit statistics (00012 = 544.28, df = 246, CFI = .92, GFI = .85, AGFI = .81, SRMR = .08, RMSEA = .071). The structure coefficients for the com­pletely standardized solution ranged from .49 to .69 for learning goals, from .50 to .61 for performance-approach goals, from .42 to .67 for performance-avoidant goals, and from .42 to .65 for work-avoidant goals. All items loaded significantly on their respective factors. Most of the four factors were also highly correlated with one another (Table 11.2). In fact, performance-approach goals and performance-avoidant goals may be correlated too highly (r = .85), suggesting that multicollinearity may be a problem. Due to the poor fit, we closely examined the modifica­tion indices. Our examination revealed that changes in parameter path­ways would result in negligible decreases in χ 2 of 2.68 or smaller. Given prior research on achievement goals and the consistently moderate to high correlations between performance­approach and performance­avoidance goals (Bong, 2004; Shih, 2005), combined with the concep­tual similarities between performance­avoidance and work­avoidant goals, we decided to conduct an exploratory factor analysis on students’ responses to the goal orientation measures. Based on prior research

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 345

and conceptual similarity, we hypothesized that a two­ (e.g., learn­ing goal factor and performance goal factor) or three­factor solution (e.g., learning goal factor, performance goal factor, and work­avoidant goal) would emerge.

Given that the selection of an appropriate number of factors was of particular importance in this analysis, we employed a series of techniques as indicators of appropriate extraction including the Kaiser–Guttman rule, parallel analysis, and scree plot analysis. The Kaiser–Guttman rule (Guttman, 1954; Kaiser, 1960) indicated the plausibility of five factors, with eigenvalues ranging from 5.420 to 1.072. However, the Kaiser–Guttman rule is known for extracting more factors than are meaningful (Linn, 1968; Tucker et al., 1969). Consequently, a parallel analysis was conducted to determine the number of plausible meaning­ful factors (Horn, 1965). This method of extracting factors has been shown to be the most accurate under diverse conditions (Zwick and Velicer, 1984). In the parallel analysis, random data were generated using a statistical package for social sciences (SPSS) programming script forwarded by Brooks (n.d.). Factors were extracted from these randomly generated data using a principle components analysis. The factors of the data collected in the study were compared to those of the random data. Those factors in the data collected from the eighth­grade sample having higher eigenvalues than those of the randomly gener­ated data were considered viable and kept for analysis. This procedure indicated that two factors should be retained to describe eighth­grade students’ goal orientation scores. These two factors explained a total of 32.750 percent of the variance in the data. The first factor accounted for 19.774 percent of the variance and the second factor accounted for 12.976 percent of the variance in the data. The selection of two factors was confirmed by an analysis of the scree plot.

Table 11.2. Correlations among goal orientations for eighth-grade students

Learning goals

Performance­approach goals

Performance­avoidant goals

Learning goals 1.00

Performance­approach goals .28 1.00

Performance­avoidant goals .08 .85** 1.00

Work­avoidant goals ­.40 .57 .67

Note. N = 255**p < .05

Epistemic beliefs and high­poverty/high­minority schools346

Direct oblimin rotation was employed because these goal orien­tations have been found to be related in prior work (e.g., Middleton and Midgley, 1997). All twenty­four of the items in the adapted goals measure had pattern coefficients greater than .35 on at least one of the factors (Table 11.3). Items having pattern coefficients above this crite­rion on more than one factor were assigned to the factor with the high­est coefficient. After applying this guideline, eighteen items loaded on factor 1 (pattern coefficients ranging from .399 to .620) and six items loaded on factor 2 (pattern coefficients ranging from .466 to .658). Two items loading on factor 1 also cross­loaded on factor 2, but not as strongly as on the first factor. These items, “I want to get others to do the school work for me” and “I want to get out of having to do school work,” loaded at .497 and .447 on factor 1 and at –.376 and –.412 on factor 2. These two items were retained on factor 1 for two reasons. First, the content of these items was similar to other items loading only on factor 1. Second, these items were negatively related to fac­tor 2. Pattern coefficients for the two­factor solution are displayed in Table 11.3.

An examination of the twenty­four items loading on each of the two factors showed that the items loading on factor 1 were representative of performance­approach, performance­avoidance, and work­avoidant goals (e.g., “I would feel successful in school if I did better than most of the other students”; “One reason I would not participate in class is to avoid looking stupid”; and “I wish I didn’t have to do school work”). Factor 2 contained items that represented learning goals (e.g., “An important reason why I do my school work is because I like to learn new things”). We refer to factor 1 as performance/work-avoidant goals and factor 2 as learning goals. The loading of performance­approach and avoidance goals on the same factor was not surprising given the high correlations between these factors in previous studies (e.g., Midgley and Urdan, 2001; Pajares et al., 2000). In addition, the inclusion of work­avoidant goals on the same factor suggested that performance and work­avoidant goals were similar in nature.

The Cronbach’s alpha for the data from all twenty­four items was .83. The reliability of the performance/work-avoidant goals factor was .86 and the reliability of the learning goals factor was .78. Composites were created for each of the goal factors by calculating the mean of the items that loaded on each factor. For example, the composite for learning goals was created by summing the six items that loaded on this factor and dividing by six. This method was employed to aid in the interpretation of the variables. That is, due to a different number of items loading on each factor, calculating the mean ensured that the composite scores

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 347

Table 11.3. Pattern coefficients for the eighth- and ninth-grade goals factors

Eighth­grade solution

Ninth­grade solution

Item F­1 F­2 F­1 F­2 F­3

62. The reason I do my school work is so my teachers don’t think I know less than others.

.620 .111 .536 –.002 .207

70. One reason I would not participate in class is to avoid looking stupid.

.608 –.263 .532 –.209 .235

57. It’s important to me that the other students in my classes think that I am good at my school work.

.586 .149 .472 .176 –.079

58. An important reason I do my school work is so that I don’t embarrass myself.

.586 .059 .595 .000 –.020

66. The reason I do my work is so others won’t think I’m dumb.

.576 –.009 .827 –.132 –.055

69. I’d like to show my teachers that I’m smarter than the other students in my class.

.569 .071 .519 .113 .232

65. I would feel successful in school if I did better than most of the other students.

.540 .141 .390 .138 .181

73. Doing better than other students in school is important to me.

.519 .242 .645 .113 –.032

56. I want to get others to do the school work for me.

.497 –.376 .288 –.217 .341

60. I want to do as little school work as possible.

.494 –.316 .075 –.087 .473

53. I would feel really good if I were the only one who could answer the teachers’ questions in class.

.490 –.004 .123 .261 .439

72. I just want to do enough school work to get by.

.483 –.072 .055 –.057 .681

74. One of my main goals is to avoid looking like I can’t do my work.

.475 .131 .572 .066 .047

64. I want to get out of having to do school work.

.447 –.412 .177 –.222 .556

54. It’s very important to me that I don’t look stupid in my classes.

.435 .301 .271 .238 .059

Epistemic beliefs and high­poverty/high­minority schools348

Eighth­grade solution

Ninth­grade solution

Item F­1 F­2 F­1 F­2 F­3

68. I want to do things as easily as possible so I won’t have to work very hard.

.410 –.142 .136 .047 .528

61. I want to do better than other students in my classes.

.409 .333 .135 .401 .321

52. I wish I didn’t have to do school work.

.399 –.266 –.183 .034 .802

55. An important reason why I do my school work is because I like to learn new things.

.049 .658 .004 .695 –.036

63. An important reason why I do my work in school is because I want to get better at it.

.025 .646 .148 .675 –.024

67. I do my school work because I’m interested in it.

–.026 .631 .159 .620 –.096

51. I like school work that I’ll learn from, even if I make a lot of mistakes.

.033 .618 –.197 .729 .103

71. An important reason I do my school work is because I enjoy it.

.010 .592 .129 .502 –.179

59. I like school work best when it really makes me think.

.097 .466 .013 .670 –.023

Note. The bold loadings for each factor indicate membership of an item on that factor.For the eighth­grade solution, F­1 = performance/work-avoidant goals, and F­2 = learning goals.For the ninth­grade solution, F­1 = performance goals, F­2 = learning goals, and F­3 = work-avoidant goals.

Table 11.3. (cont.)

for each of the factors were on the same scale (i.e., 0 to 9). As shown in Table 11.4, the correlation between the composite scores of performance/work-avoidant goals and learning goals was not statistically significant (r = .014, p = .821).

Ninth-grade goal structure. Analysis procedures for the ninth­grade data paralleled those used to examine the eighth­grade data. The confirmatory factor analysis revealed that fit of the four­ factor model was not good, mirroring the results found for the eighth­grade data (00012 = 699.62, df = 246, CFI = .90, GFI = .81, AGFI = .77,

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 349

SRMR = .077, RMSEA = .086). The structure coefficients for the com­pletely standardized solution in this model ranged from .60 to .74 for learning goals, from .48 to .70 for performance-approach goals, from .38 to .72 for performance-avoidant goals, and from .52 to .71 for work-avoidant goals. Once again, all items loaded significantly on their respective fac­tors. In addition, the four resulting factors exhibited very high intercor­relations (Table 11.5). While it appears that the structure coefficients in this model may be better than those of the eighth­grade model, multicollinearity may be more of a problem. Specifically, performance-approach goals and performance-avoidant goals were correlated at .90. In addition to the poor fit and issues of multicollinearity, the modification indices revealed that changes in parameter pathways would result in negligible decreases in χ 2 of 1.75 or smaller. Based on these results, we decided to conduct an exploratory factor analysis similar to the one conducted with the eighth­grade data.

Table 11.4. Correlations among epistemic beliefs, goal orientations, and GPA for eighth-grade students

Variable M SD 1 2 3 4

1. Beliefs about learning 6.00 1.49 1.00

2. Learning goals 5.73 1.86 .59** 1.00

3. Performance/work avoidant goals

4.56 1.60 .11 .01 1.00

4. Composite GPA 2.13 .82 .28** .24** –.12 1.00

Note. N = 255**p < .05

Table 11.5. Correlations among goal orientations for ninth-grade students

Learning goals

Performance­approach goals

Performance­avoidant goals

Learning goals 1.00

Performance­approach goals .47 1.00

Performance­avoidant goals .23 .90** 1.00

Work­avoidant goals –.15 .57 .64**

Note. N = 195**p < .05

Epistemic beliefs and high­poverty/high­minority schools350

Similar procedures were employed to analyze the goal data from the ninth­grade participants. The Kaiser–Guttman rule revealed five fac­tors with eigenvalues greater than one, ranging from 6.145 to 1.116. A parallel analysis (Horn, 1965) indicated that three factors should be retained and this was confirmed by the scree plot. These three fac­tors explained 40.667 percent of the variance, with factor 1 explain­ing 23.161 percent, factor 2 explaining 13.441 percent, and factor 3 explaining 4.064 percent of the variance in the data (Table 11.3).

Using direct oblimin rotation, three factors were extracted and the pattern coefficients were examined. A criterion of .35 was again used to determine the assignment of individual items to specific factors. Of the twenty­four items, twenty­two had pattern coefficients greater than .35 on at least one of the factors, with nine items loading on factor 1 (coefficients ranging from .390 to .827), seven items loading on factor 2 (.401 to .729), and six items loading on factor 3 (.439 to .802). Pattern coefficients for the three­factor solution are presented in Table 11.3.

An examination of the twenty­two items loading on each of the three factors revealed that students’ goal structures are more differ­entiated by the ninth­graders than they were by the eighth­graders. The items loading on factor 1 measured performance­approach and performance­avoidance orientations, whereas factor 2 contained items representative of a learning goal orientation (Ames and Archer, 1988). One item measuring a performance­approach orientation (i.e., “I want to do better than other students in my classes”) also loaded on factor 2. Items loading on factor 3 generally represented a work­avoidant orientation (Meece et al., 1988) with the exception of one item which evidenced a performance­approach orientation (i.e., “I would feel really good if I were the only one who could answer the teachers’ questions in class”). The aforementioned item represented a performance­approach orientation, and was not conceptually similar to the other items loading on factor 3. As such, both items not load­ing cleanly on the conceptualized factor were eliminated from further analysis, leaving factor 2 with six items and factor 3 with five items. We refer to factor 1 as performance goals, factor 2 as learning goals, and factor 3 as work-avoidant goals.

The Cronbach’s alpha for scores from the final twenty items load­ing on the three factors was .84. Reliabilities for scores from each of the three factors was .84 (performance goals), .83 (learning goals), and .77 (work-avoidant goals). As shown in Table 11.6, correlations among each of the three factors were also studied using composite scores. Performance goals showed a positive relationship with both learning goals (r = .245, p = .001) and work-avoidant goals (r = .488, p < .0001). Learning

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 351

goals and work-avoidant goals were not statistically significantly related (r = –.088, p = .244).

Our results indicated that there was an increase in the differentia­tion of our students’ goal orientations from the eighth­grade to the ninth­grade and that the structure of the eighth­grade students’ goals differed from previous investigations. Specifically, the eighth­grade students did not differentiate between the non­learning goals (i.e., performance­approach, performance­avoidance, and work­avoidant) but, by ninth­grade, students distinguished between learning goals, performance goals, and work­avoidant goals.

The difference between our findings and previous investigations may be attributable to the composition of our sample. That is, our par­ticipants were all students from a high­poverty, high­minority school. Further, 90 percent of our eighth­grade sample consisted of minority students. Other investigations have either applied the factor structure identified in previous studies (e.g., Guttman, 2006) or examined the structure of students’ goals with a middle­class sample or heterogene­ous samples that included both minority and non­minority students from a range of economic backgrounds (e.g., Middleton and Midgley, 1997; Midgley et al., 1996). Additionally, our measures included both performance­avoidance and work­avoidant goal items, which has not been done in previous investigations.

The lack of differentiation of the non­learning goals in the eighth­grade students indicates that students responded to the items similarly. Perhaps the distinctions indicated in the items are not meaningful to the students at this age or students possess similar levels of performance­approach,

Table 11.6. Correlations among epistemic beliefs, goal orientations, and GPA for ninth-grade students

Variable M SD 1 2 3 4 5 6

1. Beliefs about Learning

5.98 1.54 1.00

2. Beliefs about Work 5.20 1.58 .60** 1.00

3. Learning Goals 5.60 1.87 .56** .51** 1.00

4. Performance Goals 4.61 1.88 .17** .26** .25** 1.00

5. Work­Avoidant Goals

4.57 1.91 .13 .14 –.02 .53** 1.00

6. Composite GPA 1.47 1.13 .10 –.05 .15 –.13 –.24** 1.00

Note. N = 195

**p < .05

Epistemic beliefs and high­poverty/high­minority schools352

performance­avoidance, and work­avoidant goals. However, students’ responses did differentiate at the ninth­grade level.

The increased differentiation may be due to developmental changes in the students as well as changes in the school environment and culture. For instance, during adolescence the peer culture becomes increasingly important, students engage in more social comparison, and they experi­ence an increased awareness of the perceptions of others. This may account for the performance goals, with both approach and avoidance relating to social comparisons and the perceptions of others, separating from work­avoidant goals.

Additionally, there are differences in the middle school and high school environments. In comparison to some of the middle school buildings, which are designed to nurture collaborative learning environ­ments (e.g., pods), students entering high school are likely to encoun­ter a larger student body with whole­class instruction, more extreme levels of individual competition (Bryk and Thum, 1989), and rigidly enforced academic ability tracking (Seidman and French, 1997). Thus, the transition can increase stress levels and promote academic undera­chievement (Reed et al., 1995). Further, there is evidence that African­American students may perform better in more communal learning environments (e.g., Hurley et al., 2005; Serpell et al., 2006). Perhaps students who previously did not differentiate between performance­ and work­avoidant goals begin to do so when presented with stressful transitions and learning environments that are incongruent with how they are accustomed to learning.

Relations among students’ epistemic beliefs, achievement goal orientations, and academic achievement

The second purpose of this study was to explore the influence of stu­dents’ epistemic beliefs and achievement goal orientations on their academic achievement. We were also interested in the extent to which goal orientations may mediate student achievement. To address this purpose, data were submitted to correlational and path analyses. Spe­cifically, we examined the relations among students’ epistemic belief factors (e.g., beliefs about learning), achievement goal factors (e.g., learn-ing goals and performance/work-avoidant goals), and composite grade point average (i.e., GPA). Composites were created for the belief fac­tors using the same procedures employed with the goal factors. That is, we calculated the mean for the items that loaded on each factor. The students’ GPAs were used as a measure of academic achievement. This score was created by averaging students’ grades in four content

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 353

areas (i.e., history, reading, mathematics, and science), where a grade of A was given a value of 4, a B = 3, a C = 2, a D = 1, and a failing grade was given a value of 0. Students’ grades were averaged across the con­tent areas for stability and variability. As before, analyses were run separately for the eighth­ and ninth­grade data. The mean GPA for the eighth­grade sample was 2.13, with a standard deviation of .823. The mean GPA for the ninth­grade sample was lower, at 1.47, with a standard deviation of 1.128.

Relations among variables for eighth-grade students. As seen in Table 11.4, several relations emerged. Specifically, the positive corre­lations between students’ beliefs about learning and both learning goals and GPA were statistically significant. Students’ beliefs about learning were not statistically significantly related to performance/work-avoidant goals. In addition, while students’ learning goals were statistically sig­nificantly related to their GPA, their performance/work-avoidant goals were not.

Relations among variables for ninth-grade students. A similar correla­tional analysis was conducted with the ninth­grade data. The relations between ninth­grade students’ epistemic beliefs, goals, and academic achievement were somewhat different from the eighth­grade data (Table 11.6). Specifically, students’ beliefs about learning were statisti­cally significantly related to their learning goals, and beliefs about work were statistically significantly related to their learning goals and perform-ance goals. Neither beliefs about learning nor beliefs about work were sta­tistically significantly related to students’ work-avoidant goals or GPA. Work-avoidant goals was the only variable related to students’ GPA. This negative relation parallels prior findings (Meece et al., 1988).

The influence of epistemic beliefs and motivation on students’ academic achievement

Finally, we examined the extent to which students’ GPA is influenced by their beliefs about the nature of knowledge and the goals with which they approach academic tasks. In addition, we hypothesized that stu­dents’ achievement goals may be influenced by their beliefs about knowledge. This hypothesis was based on Dweck and Leggett’s propo­sition (1988) that students’ goals are influenced by their belief systems and the finding by Qian and Burrus (1996) in which two of Schom­mer’s epistemic factors were significant predictors of students’ goal orientations. Bråten and Strømsø (2004) also recently suggested that students’ beliefs may be an antecedent to their goal orientations, not­ing the beliefs about knowledge “may function as implicit theories that

Epistemic beliefs and high­poverty/high­minority schools354

can give rise to personal goals for learning” (p. 374). We agree with this speculation. Students’ beliefs about learning and knowledge, in gen­eral, may be connected to their theories about their own intelligence. To more closely examine those relations, we submitted students’ data to path analysis. The hypothesized model is displayed in Figure 11.1.

We chose to employ path analysis to address this research question because it offers at least two benefits over linear regression modeling. First, path analysis estimates the error variance associated with each variable. Second, path analysis provides a means to test the overall fit of the proposed model. Although numerous fit indices are discussed in the literature, we chose indices that are bounded by zero and one. As in the previous confirmatory factor analyses, Hu and Bentler’s (1998) criteria for assessing goodness of fit were used. To obtain indices of fit, we used EQS Version 5.7b, a program designed specifically to test path models and structural equation models (Bentler, 1998).

In addition to assessing the fit of our model, we were concerned with the statistical significance of specific paths. Maximum likelihood (ML) estimation was used to estimate these parameters because this proce­dure is robust for multivariate normal data (Bollen and Long, 1993). In EQS, the test statistic resulting from ML estimation functions as a Z statistic (Byrne, 1994). Thus, to conclude that a parameter estimate is significantly different than zero, the test statistic would need to be greater than + 1.96 at an alpha level of .05.

Eighth-grade model. For the eighth­grade students, we tested a model incorporating one epistemic belief factor, two goal factors, and stu­dents’ GPA. Specifically, beliefs about learning were hypothesized to predict learning goals and performance/work-avoidant goals. Direct paths were also included from beliefs about learning to students’ GPA and from the goal factors to students’ GPA.

Epistemic beliefs

Achievement goals

Academic achievement

Figure 11.1: Hypothesized model for the relationship among students’ epistemic beliefs, achievement goals, and academic achievement

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 355

As indicated by the fit indices in Figure 11.2, this model provided good fit of these data. However, not all of our hypothesized paths were statistically significant. Instead, the results of this analysis indicated that students’ beliefs about learning have a significant, positive direct effect on both their learning goals and GPA, but not on students’ per-formance/work-avoidant goals. Students’ performance/work-avoidant goals had a negative direct path to composite GPA. This finding implies that students’ orientations to outperform others, avoid failure, or sim­ply avoid work ultimately result in lower GPA. Previous investigations also identified such relations (e.g., Leonardi and Gialamas, 2002; Middleton and Midgley, 1997). These findings suggested that epis­temic beliefs serve as substantive predictors of GPA and an overall goal to learn content.

Ninth-grade model. Next, the hypothesized model was assessed using the ninth­grade data. Due to the different factor structures for the ninth­grade students’ beliefs about knowledge and goal orientations, this model included additional belief and goal factors (i.e., beliefs about work, performance goals, and work-avoidant goals). Given the relations between the belief factors, we allowed students’ beliefs about learning and beliefs about work to covary. We also allowed the error terms of the goal factors to covary to account for potential relations between the different goal orientations. The resulting model fit these data well (Figure 11.3), and several of the hypothesized paths were significant.

Beliefs aboutlearning

Performance/Work-avoidant

goals

Learning goals

Composite GPA

.229**

– .146*

.585*

*

CFI = 1.000GFI = .998AGFI = .981SRMR = .016RMSEA = .000

χ2 = .970df = 1

Figure 11.2: Eighth­grade model with epistemic beliefs, two­factor goals, and students’ GPA

Epistemic beliefs and high­poverty/high­minority schools356

As with the eighth­grade data, there was a significant, positive path from students’ beliefs about learning to their learning goals. Students’ beliefs about work significantly positively predicted both learning goals and performance goals. These relations support the possible influence of students’ perceptions of learning and the role of effort and work on students’ motivation. Figure 11.3 illustrates significant negative paths between students’ beliefs about work and their GPA as well as between work-avoidant goals and GPA. However, the other variables were not statistically significantly related to students’ academic performance.

Overall, the results of these path analyses suggested that students’ epistemic beliefs and achievement goals play important roles in the aca­demic achievement of students enrolled in high­poverty, high­minority schools. Further, the nature of the relations among students’ beliefs, goals, and achievement change with students’ age and transition from middle school to high school. Specifically, students’ epistemic beliefs had a positive effect on the academic achievement of the eighth­grade students, underscoring the importance of fostering these beliefs. The negative relation between eighth­grade students’ performance/work-avoidant goals implies that students’ orientations toward outperform­ing others, avoiding failure, and avoiding work are detrimental to their learning and academic performance.

Beliefs aboutlearning

Beliefs aboutwork

Performancegoals

Learning goals

Composite GPA

.403**

– .214*

.264*

.235**

.604**

CFI = .990GFI = .994AGFI = .866SRMR = .029RMSEA = .117

χ2 = 3.418df = 1

Work-avoidantgoals

– .189 **

Figure 11.3: Ninth­grade model with epistemic beliefs, three­factor goals, and students’ GPA

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 357

At ninth grade, there is a different pattern in the relations between the variables. For example, beliefs about learning no longer significantly predicted students’ academic performance. Instead, beliefs about work are a significant negative predictor of students’ GPA. Consequently, students who believe that the acquisition of knowledge requires effort and hard work tend to have lower levels of academic performance. This finding is possibly attributable to the low variation in students’ GPA.

In addition, ninth­grade students’ work-avoidant goals were found to have a negative influence on the achievement. This finding suggests that work-avoidant goals are powerful predictors of the achievement of students enrolled in high poverty, high­minority schools. Given the lack of attention to this construct in recent studies, additional research on work­avoidant goals and the role they play in the academic development of students enrolled in high­poverty, high­minority schools is needed. Finally, learning goals and performance goals did not relate to students’ academic performance. This lack of significance suggests that these goals do not play an important role in the achievement of inner­city stu­dents. Consequently, the conceptual distinction between performance­approach and performance­avoidance goals is of little consequence.

Conclusions and implications

The purpose of this study was to examine the epistemic beliefs, achieve­ment goals, and academic achievement of students enrolled in high­poverty, high­minority schools. Specifically, we were interested in the dimensionality of students’ goal orientation data, as well as the relations among epistemic beliefs, goal orientations, and achievement of eighth­ and ninth­grade students enrolled in high­poverty, high­minority schools. However, we must first acknowledge several limitations with this investigation.

The limitations relate primarily to our sample. We collected data from 450 students enrolled in high­minority, high­poverty schools but we did not have data from a comparison group. Thus, it is difficult to conclude whether our results are generalizable to other adolescents who do not attend similar schools. Instead, we must simply compare our findings to those of other studies, many of which did not present complete information about their participants. In addition, the size of our sample presented certain limitations. That is, we had data from 255 eighth­grade students and 195 ninth­grade students. These numbers prevented us from employing more sophisticated statistical methods such as structural equation modeling. Finally, although this study set

Epistemic beliefs and high­poverty/high­minority schools358

out to investigate the influence of students’ beliefs and goals on their academic achievement, students’ achievement was not particularly high, limiting the variability in our data.

Despite these limitations, the findings of this investigation offer inter­esting additions to the existing research literature as well as to educa­tional practice. Due to concerns about the generalizability of previous investigations, we felt that it was important to examine the structure of goal orientations for an underrepresented population, namely those of students learning in high poverty schools. Our findings indicated that some of the distinctions made between the different types of goal orien­tations are not meaningful for this population. That is, whereas a learn­ing goal factor emerged for both the eighth­ and ninth­grade samples, performance and work­avoidant goals were not well differentiated at the eighth­grade level. As discussed in the results and discussion section, this difference between the eighth­ and ninth­grade students may be attributable to developmental changes as well as changes in the school environment and culture that place less emphasis on collaborative and communal learning and more on competition and social comparison. Further, peer culture and students’ perceptions of the value of educa­tion may contribute to the emerging distinctions between performance and work­avoidant goals.

Similarly, although separate performance and work­avoidant fac­tors emerged for the ninth­grade students, scores on this measure did not differentiate between performance­approach and performance­ avoidance goals. Thus, our results suggested that more attention needs to be devoted to exploring how various constructs are manifested and measured in diverse populations, as well as how students’ beliefs and goals develop over time. Specifically, additional research is needed to determine if the approach­avoidance distinction is meaningful for non­white, middle­class populations at varying ages. Previous investigations identifying specific performance goals have often utilized a Caucasian, middle­class sample or a heterogeneous sample with a mix of ethnici­ties and/or socio­economic backgrounds. Thus, future research should first determine if the distinction emerges in homogeneous samples of non­Caucasian students. Subsequent research can then determine how performance­approach and performance­avoidance goals are associ­ated with adaptive or maladaptive learning outcomes for students from varied backgrounds as well as explore how students’ goal orientations change over time.

Our study also provides some insight into relations among students’ epistemic beliefs, goal orientation, and academic achievement. For instance, we found evidence of the hypothesized relations between

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 359

students’ beliefs about knowledge and learning and their achievement goal orientations (e.g., Bråten and Strømsø, 2004; Buehl, 2003). That is, at eighth­ and ninth­grade there were positive direct paths from students’ beliefs about learning to their learning goals. At the ninth­grade, there were also positive direct paths from students’ beliefs about the need for hard work and effort to learning goals and performance goals. These relations suggest students’ beliefs about knowledge and learning may serve as antecedents to their goal orientations (Bråten and Strømsø, 2004). Additional research is needed to explain the reason for this relation. That is, we would need to assess students’ general beliefs about knowledge and learning, their implicit theories of their own intel­ligence, as well as their goal orientations to determine if the connec­tion between epistemic beliefs and goals is mediated by students’ beliefs about themselves. This string of influences is reminiscent of work by Lent et al. (2005) regarding self­efficacy beliefs and their effect upon goals. Interestingly enough, in the same study, African­American stu­dents reported stronger self­efficacy and educational goals than Cau­casians, but earlier results indicated that efficacy only impacted grade performance in one domain (Lent et al., 1993). Thus the relation­ship, which can vary by ethnic group and content area, is tangentially dependent upon a variety of factors.

Further, our investigation revealed significant relations between stu­dents’ beliefs and their academic performance, what some have termed epistemic beliefs about learning. In the eighth­grade data, beliefs about learning were positively related to performance. Consequently, it may be beneficial for teachers to support the development of students’ beliefs about knowledge. For example, it may be useful for students to actively engage in discussions about learning and what it means to know. As Schommer­Aikins et al. (2005) suggested, if students con­ceive knowledge as easily acquired and learning as a quick process, they may become discouraged easily with demanding assignments or chal­lenges in life. At the ninth­grade level, we identified a negative rela­tion between students’ beliefs about the need for hard work in order to acquire knowledge and their academic performance. This finding was unexpected and additional exploration is needed to fully understand this effect.

The apparent resistance to challenging tasks may be a predictable response to feelings of inadequacy. Indeed, students who do not acquire phonetic spelling rules or continuity and fluency in reading while they are in elementary school (Kozol, 2005) may still be lacking those skills in high school. When faced with demanding writing, reading, and eval­uative assignments that are beyond their current levels of preparedness,

Epistemic beliefs and high­poverty/high­minority schools360

students are evidently more likely to formulate compensating strate­gies that are woefully insufficient. Thus, although students recognize the need for hard work, they may doubt their ability to engage in the necessary tasks (e.g., self­efficacy). However, this presumes that stu­dents are cognizant of the subtask necessary to perform well. Perhaps students realize that work is important to success but cannot analyze and conceptualize the specific work, or tasks, that are necessary. An alternative explanation may relate to students’ beliefs about the value of education. Steinberg et al. (1992) found that compared to European­American students, African­American students did not view education as essential for future success in life. If the students in our sample do not value academic success, they may be less likely to engage in the behaviors needed for success (e.g., working hard), even if they recog­nize the connection between hard work and academic success. Also, the recognition of the need for hard work may act as a deterrent if students are concerned about the perceptions of peers.

With respect to goal orientations, the results of our investigation sug­gest that learning goals are not directly related to the achievement of students enrolled in high­poverty, high­minority schools. Thus, future research needs to focus on motivational constructs that might be more relevant for this population (e.g., value of education, outcome expect­ancy, self­efficacy). Additionally, it may be beneficial to determine why students who express a desire to learn and master the material do not do so. This may be a reflection of students lacking the skills needed to suc­ceed, constraints in the school environment (e.g., lack of resources), or competing priorities in the students’ lives. Understanding this relation, or lack of relation, is needed prior to attempting to increase students’ goals in the absence of other factors.

It is also important to examine how beliefs about learning and learn­ing goals can be sustained or strengthened. One suggestion obviously relates to the types of knowledge that students have been encouraged to acquire. It is impossible for students whose teachers focus on their stu­dents learning how to fill out insurance forms and acquire good working habits to compete with peers who are taught to think critically, crea­tively, and learn scientific principles (Oakes, 2005). Similarly, students who are routinely taught to copy directions for doing an experiment (Finn, 1999) are not being prepared to logically design experiments. Within such diverse contexts, how learning goals and beliefs about learning can influence achievement.

Based on the findings of this study, students’ learning goals are not related to achievement as indicated in previous investigations. Further, work­avoidant goals have a significant negative impact on students’

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 361

achievement. In essence, our participants’ performance outcomes were based on external motivators (e.g., receiving the approval of others and avoiding punishment or work), not more internal motivators (e.g., a desire to learn and master the material). Such relations were exac­erbated by the transition from eighth­grade to ninth­grade. That is, although the mean levels of students’ learning, performance, and work­avoidant goals appear comparable, based on descriptive comparisons of the means for different factor structures across the eighth­ and ninth­grades, the negative relation between the work­avoidant related factors and academic achievement is stronger in ninth­grade than eighth­grade. Additional research must seek to further understand the antecedents to students’ work­avoidant goals as well as how to combat them once they are formed. Attending to how and why this construct develops and manifests itself in students may provide an advantageous route to improving students’ academic achievement. In terms of current prac­tices, teachers need to be cognizant of students’ goals, particularly their work­avoidant goals, and foster adaptive work habits. Further, it may be beneficial for teachers to address the value of education in general, and relative to specific content areas, as well as to address students’ efficacy and expectancy outcome beliefs. Students need to believe that they can succeed and that their efforts can lead to the desired outcome (i.e., not be hindered by discrimination or lack of opportunity).

Perhaps the most substantive outcome of our research, however, is that findings from prior research in epistemology and motivation are not generalizable to the students participating in our study. Indeed, as most would posit, students living and going to school in high­poverty areas experience a very different existence than students living in middle­class, suburban neighborhoods. Students in high­poverty schools face ongo­ing issues related to physical safety, transiency, teacher turnover, poor instruction, and lack of resources (e.g., Kozol, 2005; Lipman, 2004; Oakes, 2005), in addition to issues they may encounter within their home environments (e.g., low­income, single­parent homes). There is a desperate need for researchers to pause and consider the applicability and generalizability of measures for different populations.

Future studies need to employ quantitative and qualitative methods to identify the beliefs and motivations that will support students’ learn­ing and achievement. Given that there may be differences in relations across ethnic or socio­economic groups (e.g., Graham, 1994), more with­in­group analyses are needed, and researchers must be sensitive to con­founding issues related to ethnicity and socio­economic status. Indeed, if students in high­poverty settings are to excel in learning, they must be motivated by an internal drive to learn and the recognition that such effort

Epistemic beliefs and high­poverty/high­minority schools362

is worthwhile. However, to foster this drive, we must first gain a deeper understanding of the beliefs and motivations of students from varied con­texts and backgrounds that is fueled by sound research methodology.

R EF ER ENCES

Alexander, P. A. (1997). Mapping the multidimensional nature of domain learning: The interplay of cognitive, motivational, and strategic forces. In M. Maehr and P. Pintrich (Eds.), Advances in motivation and achievement (10, 213–50). Greenwich, CT: JAI Press.

Ames, C. (1992). Classrooms: Goals, structures, and student motivation. Jour-nal of Educational Psychology, 84, 261–71.

Ames, C. and Archer, J. (1988). Achievement goals in the classroom: Students’ learning strategies and motivation processes. Journal of Educational Psy-chology, 80, 260–67.

Baxter Magolda, M. B. (1992). Knowing and reasoning in college: Gender-related patterns in students’ intellectual development. San Francisco: Jossey­Bass.

Beck, L., Kratzer, C., and Isken, J. (1997). Caring for transient students in one urban elementary school. Journal for Just and Caring Education 3(3), 343–69.

Bentler, P. M. (1998). EQS 5.7b. Encino, CA: Multivariate Software, Inc.Blumenfeld, P. C. (1992). Classroom learning and motivation: Clarifying and

expanding goal theory. Journal of Educational Psychology, 84, 272–81.Bollen, K. A. and Long, J. S. (1993). Testing structural equation models. Newbury

Park, CA: Sage Publications.Bong, M. (2004). Academic motivation in self­efficacy, task value, achieve­

ment goal orientations, and attributional beliefs. Journal of Educational Research, 97(6), 287–97.

Bråten, I. and Strømsø, H. I. (2004). Epistemological beliefs and implicit theo­ries of intelligence as predictors of achievement goals. Contemporary Edu-cational Psychology, 29, 371–88.

Brooks, G. (n.d.). Parallel analysis program. Retrieved October 13, 2004, from http://oak.cats.ohiou.edu/~brooksg/edre/parallel_analysis.sps.

Bryk, A. and Thum, Y. (1989). The effects of high school organization on dropping out: An exploratory investigation. American Educational Research Journal, 26, 353–83.

Buehl, M. M. (2003). At the crossroads: Exploring the intersection of epis­temological beliefs, motivation, and culture. In H. Fives (Chair), Inter-nationalizing the study of epistemology, goal orientations, and self-efficacy. Symposium presented at the annual meeting of the American Educational Research Association, Chicago (April).

Buehl, M. M., Alexander, P. A., and Murphy, P. K. (2002). Beliefs about schooled knowledge: Domain specific or domain general? Contemporary Educational Psychology, 27, 415–49.

Byrne, B. M. (1994). Structural equation modeling with EQS/Windows: Basic concepts, applications, and programming. Thousand Oaks, CA: Sage Publications.

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 363

Conley, A. M., Pintrich, P. R., Vekiri, I., and Harrison, D. (2004). Changes in epistemological beliefs in elementary science students. Contemporary Edu-cational Psychology, 29, 186–204.

Dweck, C. (1986). Motivational processes affecting learning. American Psy-chologist, 41, 1040–8.

Dweck, C. S. and Legget, E. L. (1988). A social­cognitive approach to motiva­tion and personality. Psychological Review, 95, 56–73.

Dweck, C. S., Tenney, Y., and Dinces, N. (1982). Implicit theories of intel­ligence as determinants of achievement goal choice. Unpublished manu­script, Cambridge, MA.

Elliot, A. J. and Harackiewicz, J. M. (1996). Approach and avoidance goals and intrinsic motivation: A mediational analysis. Journal of Personality and Social Psychology, 70, 461–75.

Finn, P. (1999). Literacy with an attitude: Educating working-class children in their own self-interest. New York: State University of New York Press.

Fordham, S. and Ogbu, J. (1986). Black students’ school success: Coping with the “burden of ‘acting white’.” Urban Review, 18, 178–204.

Gehlbach, H. (2006). How changes in students’ goal orientations relate to out­comes in social studies. The Journal of Educational Research, 99, 358–70.

Graham, S. (1994). Motivation in African Americans. Review of Educational Research, 64, 55–117.

Guttman, L. (1954). Some necessary conditions for common factor analysis. Psychometrika, 19, 149–61.

Guttman, L. M. (2006). How student and parent goal orientations and classroom goal structures influence the math achievement of African Americans during the high school transition. Contemporary Educational Psychology, 31, 44–63.

Harackiewicz, J. M., Durik, A. M., and Barron, K. E. (2005). Multiple goals, optimal motivation, and the development of interest. In J. P. Forgas, K. D. Williams, and S. M. Laham (Eds.), Social motivation: Conscious and uncon-scious processes (21–39), Cambridge: Cambridge University Press.

Hofer, B. K. (1999). Instructional context in the college mathematics class­room: Epistemological beliefs and student motivation. Journal of Staff, Program, and Organization Development, 16, 73–82.

(2000). Dimensionality and disciplinary differences in personal epistemol­ogy. Contemporary Educational Psychology, 25, 378–405.

Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learn­ing. Review of Educational Research, 67, 88–140.

Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 30, 179–85.

Hu, L. and Bentler, P. M. (1998). Fit indices in covariance structure modeling: Sensitivity to underparameterized model misspecification. Psy-chological Methods, 3, 424–53.

Hurley, E., Boykin, A., and Allen, B. (2005). Communal versus individual learning of a math­estimation task: African American children and the culture of learning contexts. Journal of Psychology: Interdisciplinary and Applied, 139(6), 513–27.

Epistemic beliefs and high­poverty/high­minority schools364

Jehng, J. J., Johnson, S. D., and Anderson, R. C. (1993). Schooling and stu­dents’ epistemological beliefs about learning. Contemporary Educational Psychology, 18, 23–35.

Kaiser, H. F. (1960). The application of electronic computers to factor analysis. Educational and Psychological Measurement, 20, 141–51.

Kardash, C. M. and Scholes, R. J. (1996). Effects of preexisting beliefs, epistemological beliefs, and need for cognition on interpretation of con­troversial issues. Journal of Educational Psychology, 88, 260–71.

King, P. M. and Kitchener, K. S. (1994). Developing reflexive judgment: Under-standing and promoting intellectual growth and critical thinking in adolescents and adults. San Francisco: Jossey­Bass.

Kozol, J. (2005). The shame of the nation: The restoration of apartheid schooling in America. New York: Crown.

Kuhn, D., Cheney, R., and Weinstock, M. (2000). The development of episte­mological understanding. Cognitive Development, 15, 309–28.

Kuhn, D. and Weinstock, M. (2002). What is epistemological thinking and why does it matter? In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

Lent, R. W., Brown, S. D., Sheu, H., Schmidt, J., Brenner, B. R., Gloster, C. S., et al. (2005). Social cognitive predictors of academic interests and goals in engineering: Utility for women and students at historically black universi­ties. Journal of Counseling Psychology, 52, 84–92.

Lent, R. W., Lopez, F. G., and Bieschke, K. J. (1993). Predicting mathematics­related choice and success behaviors: Test of an expanded social cognitive model. Journal of Vocational Behavior, 42, 223–36.

Leonardi, A. and Gialamas, V. (2002). Implicit theories, goal orientations, and perceived competence: Impact on students’ achievement behavior. Psy-chology in the Schools, 39, 279–91.

Linn, R. L. (1968). A Monte Carlo approach to the number of factors problem. Psychometrika, 38, 37–72.

Lipman, P. (2004). High stakes education: Inequality, global, and urban school reform. New York: Routledge Falmer.

Long, J. F., Monoi, S., Harper, B., Knoblauch, D., and Murphy, P. K. (2007). Academic motivation and achievement among urban adolescents. Urban Education, 42, 196–222.

Meece, J. L., Blumenfeld, P. C., and Hoyle, R. (1988). Students’ goal orienta­tions and cognitive engagement in classroom activities. Journal of Educa-tional Psychology, 80, 514–23.

Meece, J. L. and Holt, K. (1993). A pattern analysis of students’ achievement goals. Journal of Educational Psychology, 85, 582–90.

Middleton, M. J. and Midgley, C. (1997). Avoiding the demonstration of lack of ability: An unexplored aspect of goal theory. Journal of Educational Psy-chology, 89, 710–18.

Midgley, C., Arunkumar, R., and Urdan, T. C. (1996). “If I don’t do well tomorrow, there’s a reason”: Predictors of adolescents’ use of academic self­handicapping strategies. Journal of Educational Psychology, 88, 423–34.

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 365

Midgley, C., Kaplan, A., Middleton, M., Maehr, M. L., Urdan, T., Anderman, L. H., et al. (1998). The development and validation of scales assessing students’ achievement goal orientations. Contemporary Educational Psy-chology, 23, 113–31.

Midgley, C. and Urdan, T. (2001). Academic self­handicapping and achieve­ment goals: A further examination. Contemporary Educational Psychology, 26, 61–75.

Murphy, P. K., Edwards, M. N., Buehl, M. M., and Zeruth, J. A. (2007). An examination of the psychometric properties of the domain-specific beliefs questionnaire (DSBQ) with urban adolescents. Journal of Experimental Edu-cation, 76, 3–25.

Niemivirta, M. (2002). Motivation and performance in context: The influence of goal orientations and instructional setting on situational appraisals and task performance. Psychologia: An International Journal of Psychology in the Orient, 45, 250–70.

Oakes, J. (2005). Keeping track: How schools structure inequality. New Haven, CT: Yale University Press.

Pajares, F., Britner, S. L., and Valiante, G. (2000). Relation between achieve­ment goals and self­beliefs of middle school students in writing and sci­ence. Contemporary Educational Psychology, 25, 406–22.

Paulsen, M. B. and Feldman, K. A. (1999). Epistemological beliefs and self­regulated learning. Journal of Staff, Program, and Organizational Develop-ment, 16(2), 83–91.

Perry, W. G. (1970). Forms of intellectual and ethical development in the college years: A scheme. New York: Holt, Rinehart and Winston.

Pintrich, P. R. (2000). An achievement goal theory perspective on issues in motivation terminology, theory, and research. Contemporary Educational Psychology, 25, 92–104.

Pintrich, P. R., Marx, R., and Boyle, R. (1993). Beyond cold conceptual change: The role of motivational beliefs and classroom contextual factors in the proc­ess of conceptual change. Review of Educational Research, 63, 167–99.

Qian, G. and Alverman, D. (1995). Role of epistemological beliefs and learned helplessness in secondary school students’ learning science concepts from text. Journal of Educational Psychology, 87, 282–92.

Qian, G. and Burrus, B. M. (1996). The role of epistemological beliefs and motivational goals in ethnically diverse high school students’ learning from science texts. In D. J. Leu et al. (Eds.), Literacies for the 21st Century (159–69). Chicago: National Reading Conference.

Reed, D., McMillan, J., and McBee, R. (1995). Defying the odds: Middle schoolers in high­risk circumstances who succeed. Middle School Journal, 27, 3–10.

Reyes, O., Gillock, K., Kobus, K., and Sanchez, B. (2000). A longitudinal examination of the transition into senior high school for adolescents from urban, low­income status, and predominantly minority backgrounds. American Journal of Community Psychology, 28, 519–44.

Schoenfeld, A. (1983). Beyond the purely cognitive: Belief systems, social cog­nitions, and metacognitions as driving forces in intellectual performance. Cognitive Science, 7, 329–63.

Epistemic beliefs and high­poverty/high­minority schools366

Schommer, M. (1990). Effects of beliefs about the nature of knowledge on comprehension. Journal of Educational Psychology, 82, 498–504.

(1993). Epistemological development and academic performance among secondary students. Journal of Educational Psychology, 85, 406–11.

Schommer, M., Clavert, C., Gariglietti, G., and Baja, A. (1997). The develop­ment of epistemological beliefs among secondary students: A longitudinal study. Journal of Educational Psychology, 89, 37–40.

Schommer, M., Crouse, A., and Rhodes, N. (1992). Epistemological beliefs and math text comprehension: Believing it is simple does not make it so. Journal of Educational Psychology, 84, 435–43.

Schommer, M. and Dunnell, P. A. (1994). A comparison of epistemological beliefs between gifted and non­gifted high school students. Roeper Press, 16, 207–10.

Schommer­Aikins, M. (2004). Explaining the epistemological belief sys­tem: Introducing the embedded systemic model and coordinated research approach. Educational Psychologist, 39, 19–29.

Schommer­Aikins, M., Duell, O. K., and Hutter, R. (2005). Epistemological beliefs, mathematical problem­solving beliefs, and academic performance of middle school students. The Elementary School Journal, 105(3), 289–304.

Schommer­Aikins, M., Mau, W., Brookhart, S., and Jutter, R. (2000). Under­standing middle students’ beliefs about knowledge and learning using a multidimensional paradigm. Journal of Educational Research, 94, 120–8.

Schraw, G., Bendixen, L., and Dunkle, M. (2002). Development and valida­tion of the epistemic belief inventory (EBI). In B. K. Hofer and P. R. Pin­trich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

Seidman, E., Allen, L., Aber, J. L., Mitchell, C., and Feinman, J. (1994). The impact of school transitions in early adolescence on the self­system and per­ceived social context of poor urban youth. Child Development, 65, 507–22.

Seidman, E. and French, S. E. (1997). Normative school transitions among urban adolescents: When, where, and how to intervene. In H. J. Walberg et al. (Eds.), Children and youth: Interdisciplinary perspectives (166–89). Thousand Oaks, CA: Sage.

Serpell, Z., Boykin, A., Madhere, S., and Nasim, A. (2006). The significance of contextual factors in African American students’ transfer of learning. Journal of Black Psychology, 32(4), 418–41.

Shih, S. (2005). Role of achievement goals in children’s learning in Taiwan. Journal of Educational Research, 98(5), 310–19.

Skaalvik, E. M. (1997). Self­enhancing and self­defeating ego orienta­tions: Relations with task and avoidance orientation, achievement, self­perceptions, and anxiety. Journal of Educational Psychology, 89, 71–81.

Steele, C. M. (1997). A threat in the air: How stereotypes shape intellectual identity and performance. American Psychologist, 52, 613–29.

Steinberg, L., Dornbusch, S., and Brown, B. (1992). Ethnic differences in ado­lescent achievement: An ecological perspective. American Psychologist, 47, 723–9.

Tanaka, A. and Yamauchi, H. (2001). A model for achievement motives, goal orientations, intrinsic interest, and academic achievement. Psychological Reports, 88, 123–35.

P. K. Murphy, M. M. Buehl, J. A. Zeruth, et al. 367

Tucker, L. R., Koopman, R. F., and Linn, L. R. (1969). Evaluation of factor analytic research procedures by means of simulated correlation matrices. Psychometrika, 34, 421–60.

Urdan, T. and Midgley, C. (2003). Changes in the perceived classroom goal structure and pattern of adaptive learning during early adolescence. Con-temporary Educational Psychology, 28, 524–51.

Wolters, C. A. (2004). Advancing achievement goal theory: Using goal struc­tures and goal orientations to predict students’ motivation, cognition, and achievement. Journal of Educational Psychology, 96, 236–50.

Zwick, W. R. and Velicer, W. F. (1984). Comparison of five results for deter­mining the number of components to retain. Psychological Bulletin, 99(3), 432–42.

368

12 Using cognitive interviewing to explore elementary and secondary school students’ epistemic and ontological cognition

Jeffrey A. Greene University of North Carolina at Chapel Hill

Judith Torney­Purta University of Maryland at College Park

Roger Azevedo University of Memphis

Jane Robertson University of North Carolina at Chapel Hill

The allure of personal epistemology research is the implicit assump­tion that in some cases students’ poor academic performance may not be due to any deficiency in skill or ability, but rather due to their naïve beliefs about knowledge and knowing (Hofer and Pintrich, 1997). Numerous researchers have asserted that naïve or less sophisticated beliefs about the nature of knowledge and knowing may cause stu­dents to memorize facts rather than construct conceptual knowledge, poorly monitor and regulate their learning, exert little effort, and neglect essential thinking skills (Hofer, 2002, 2004a, 2004b). These students act as mere receivers of information (Belenky et al., 1986) and not active constructors of knowledge who appropriately question authority figures or other sources of knowledge (Baxter Magolda, 2004; King and Kitchener, 1994; Perry, 1970). However, accord­ing to personal epistemology researchers, if these students could be shown that knowledge is complex and dynamic, they would see the need for critical thinking skills to evaluate the justifications for knowl­edge claims and be able to construct complex and coherent knowledge whose warrants could be both appropriately scrutinized when neces­sary and utilized in a convincing manner on tests, papers, and other academic assignments (Kuhn and Weinstock, 2002). Such sophistica­tion is often described as a prerequisite for success in higher educa­tion (Boyer, 1987; Sullivan and Rosin, 2008). Yet, the research on personal epistemology is not clear whether elementary or secondary school students are likely to have or need such sophisticated beliefs regarding the nature of knowledge and knowing.

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 369

In fact, there are legitimate questions regarding whether current models and measures of personal epistemology reveal much at all about the beliefs and cognition of students younger than college age. Numerous conceptual models have built upon the work of Perry (1970), claiming that elementary and secondary students, barring rare exceptions, hold naïve beliefs. These claims are challenged by work from developmental psychology, particularly theory of mind research (Chandler et al., 2002; Flavell, 2004). Other models of personal epistemology (Hofer and Pintrich, 1997; Schommer, 1990) allow for the possibility of sophisticated epistemic cognition prior to college age, but the empirical support for these models suffers from psychometric problems including poor measurement valid­ity and reliability (Buehl, 2008; Clarebout et al., 2001). Therefore, given these conceptual and methodological concerns, we argue that the field would benefit from both a re­evaluation of its conceptual foundations as well as more exploratory, qualitative investigations regarding the nature of elementary and secondary students’ beliefs about knowledge and knowing.

In this chapter we briefly review the literature on personal episte­mology and the gaps in research on elementary and secondary stu­dents, outlining both conceptual and measurement­based concerns. Next we introduce our model of epistemic and ontological cognition, which is informed by other conceptual models, philosophical episte­mology, findings from developmental psychology, and Hofer’s (2001) call for integration within the field. From there we advance cogni­tive interviewing (Karabenick et al., 2007; Willis, 2005) as a means of gathering more information regarding the phenomena of epistemic and ontological cognition as well as the suitability of questionnaires designed to measure these constructs. Cognitive interviewing has only recently been adapted for use with children, and to our knowledge it has never been used to investigate measures of epistemic and onto­logical cognition with pre­college age students. The results of our own cognitive interviewing study demonstrate the importance of vetting questionnaire items with participants, as we found numerous confu­sions regarding academic domains, the meaning of Likert response scales, and students’ interpretations of words such as “truth” and “believe.” Our findings also shed light upon deeper conceptual issues such as elementary and secondary school students’ ability to think in a sophisticated manner about ill­structured domains and their hierar­chical views of authority figures. We end the chapter with a discussion of the conceptual and educational implications of these findings and our model.

Using cognitive interviewing370

Personal epistemology models

As has been well­described elsewhere, conceptual models of personal epistemology can be grouped into those that focus more on developmen­tal stages or phases and those that emphasize systems of beliefs on which individuals can vary asynchronously (Hofer and Pintrich, 1997). The developmental models (Baxter Magolda, 2004; Chandler et al., 2002; King and Kitchener, 1994, 2004; Kuhn et al., 2000; Perry, 1970) outline qualitatively different positions, levels, or stages of development within personal epistemology, and posit a developmental progression from positions that are less sophisticated to those that are more sophisticated regarding the nature of knowledge and knowing. While the names and details of the positions differ across models, they have significant over­lap, and Kuhn and colleagues’ absolutist, multiplist, and evaluativist levels provide a sufficient summary. The absolutist sees the world, and thus knowledge, as objective, with the justification for knowledge claims coming from others such as authority figures. The multiplist holds that knowledge is subjective, with justification seen as personal, and immune to outside scrutiny or question. Finally, evaluativists are posited to view knowledge as having both objective and subjective aspects, with justi­fication for knowing established using critical thinking and evaluative criteria such as rationalism and the investigation of sources.

Developmental models of personal epistemology focus upon posi­tions or stages, but characterize those positions based upon differences across numerous categories that resemble dimensions. For example, Kuhn and colleagues use differences in individuals’ beliefs regard­ing reality, knowledge, critical thinking, and assertions as indicators of the qualitatively different levels. Yet, their focus is upon the levels, and not upon how individuals’ dimensional beliefs cohere and change systematically over time, or how people’s beliefs along these dimen­sions may predict academic outcomes. Systems of beliefs models, such as those by Schommer (1990; Schommer­Aikins, 2004) and Hofer and Pintrich (Hofer, 2004b; Hofer and Pintrich, 1997) do focus upon these dimensions, how they change, and their predictive utility in terms of various outcomes of interest. While the names and foci of the dimen­sions differ across models, in general they can all be characterized as pertaining to either the nature of knowledge or the nature of knowing (Hofer and Pintrich, 1997).1 The former concern whether individuals

1 Schommer­Aikins’ (2004) “nature of learning” dimensions have generally been viewed as important covariates in terms of understanding the influence of student beliefs upon academic performance, but not truly epistemological in nature (see Hofer and Pintrich, 1997).

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 371

see knowledge as simple and certain or complex and dynamic. The latter address beliefs regarding whether knowledge claims can be justi­fied, and the acceptable sources of that justification, such as authority figures.

Whether emphasizing developmental positions or a system of beliefs, there are a number of important similarities across these personal epis­temology models, including a common focus upon beliefs regarding simple and certain knowledge and the role of authority figures in justifi­cation (Bendixen and Rule, 2004; Hofer and Pintrich, 1997; Muis et al., 2006). Yet, the field of personal epistemology continues to be hampered by debates regarding construct definition (Hofer, 2004a), the degree of domain­generality or specificity of the phenomena (Alexander, 2006; Buehl et al., 2002; Hofer, 2006), and the poor psychometric quality of various quantitative measurement instruments (Buehl, 2008). These concerns certainly influence how the field conceives of college­aged individuals, but they are even more problematic in terms of understand­ing elementary and secondary students’ beliefs regarding the nature of knowledge and knowing.

Elementary and secondary school students’ personal epistemology

Educational researchers’ understanding of elementary and secondary students’ beliefs about the nature of knowledge and knowing is meager (Hofer, 2008). Perry, arguably the founder of the field, is often char­acterized as asserting that individuals hold a dualistic, or what Kuhn would call an absolutist, position until the college years. However, Perry is quite clear that his data reflect only college students’ experiences, and that any information regarding beliefs prior to college is based upon his respondents’ retrospection. Nonetheless, he does suggest that it would be highly unlikely for individuals to move beyond dualism prior to col­lege because he characterizes sustained experience in higher education as the principal catalyst of this intellectual development. Indeed, Perry (1999) uses a Garden of Eden analogy, likening college to the “ser­pent” (p. 67) that exposes humanity to the truth about knowledge and the necessity of judgment. From this characterization, and analyses based upon retrospective data, Perry laid the groundwork for a gen­eration of research into personal epistemology that was predisposed to look for development only after matriculation to higher education. For example, while Baxter Magolda’s (2004) work only included college students, two­thirds of her initial freshmen sample was found to be in the lowest position in her model, implying that individuals younger

Using cognitive interviewing372

than college age almost certainly held these naïve beliefs. King and Kitchener (2004) have reported that across twenty­five studies and 1,500 respondents ranging from high school to graduate students “the high school students consistently evidenced the assumptions associated with their most naïve position, prereflective thinking” (p. 15).

Chandler and colleagues (2002) have pointed out that findings from theory of mind research (see Flavell, 2004, for a review) conflict with those in personal epistemology research, indicating that children as young as seven can display sophisticated beliefs regarding the nature of knowledge and knowing, depending upon their familiarity with the knowledge domain in question. For example, when considering argu­ments regarding the appropriate legal driving age, adolescents can dis­play relativistic thinking, but they still think in an absolutist way about areas such as science (Boyes and Chandler, 1992). An explanation for these differing findings between theory of mind and developmental personal epistemology research is that the former allow for beliefs about knowledge and knowing to vary across domains, whereas the latter tend to take a domain­general view. Under this latter view, individuals are characterized according to their least sophisticated beliefs. If individuals display evaluativist thinking in every domain except physics, in which they think in an absolutist manner, they would be labeled as absolutists by personal epistemology researchers holding a domain­general per­spective. Thus, the classification system used by researchers working under a domain­general model make it likely that pre­college individu­als will be seen as naïve (Chandler et al., 2002; Hallet et al., 2002). One of the few domain­specific developmental models of personal episte­mology was advanced by Kuhn and colleagues, and they did find evi­dence of evaluativist thinking in children as young as ten­years­old, but only in terms of aesthetics. Kuhn and colleagues reported that an abso­lutist view of academic domains was the norm for participants younger than college­age, causing Chandler and colleagues (2002) to question whether their measure was accurately capturing epistemic cognition.

Numerous researchers working with multidimensional models of epistemological beliefs have investigated students younger than col­lege­age, expecting to find relevant and informative variance in terms of their beliefs about the nature of knowledge and knowing (Conley et al., 2004; Kuhn et al., 2000; Qian and Alverman, 1995; Schommer, 1990, 1993; Schommer et al., 1992; Schommer­Aikins et al., 2000; Schommer­Aikins and Easter, 2006). While Schommer­Aikins is one of the most prolific and influential researchers in the area of epistemo­logical beliefs, the psychometric qualities of her epistemological ques­tionnaire (EQ) have been called into question (Buehl, 2008; Clarebout

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 373

et al., 2001). Unfortunately, analyses of her EQ with middle and high school student samples have revealed varying three and four factor solutions, as well as low reliabilities for those factor scores, with the only consistent finding being an inverse relationship between belief in quick learning and academic performance (e.g., Neber and Schom­mer­Aikins, 2002; Qian and Alverman, 1995; Schommer et al., 1997; Schommer­Aikins et al., 2000). To be fair, researchers have been criti­cal of all of the quantitative measures used in personal epistemology research, listing concerns regarding what beliefs should be measured, the low predictive validity between these beliefs and outcomes of inter­est, and the use of homogeneous samples leading to restriction of range issues (Schraw and Olafson, 2008; Wood and Kardash, 2002). In a wide­ranging review, Buehl (2008) demonstrated that the issues regarding the psychometric qualities of measures of personal epistemol­ogy apply to instruments used with elementary and secondary school students just as they do to college and graduate students.

Overall, Hofer (2008) has said “there is too little research in child­hood and adolescence to ascertain what happens during these rather large formative periods, and what research we do have suggests that stages once expected to appear initially in college may actually happen earlier” (p. 6). Therefore, a new model is needed that can integrate both developmental and systems of beliefs models of personal epistemology, as well as align with findings from developmental psychology regarding elementary and secondary school students.

Our conceptual model of epistemic and ontological cognition

In a previous publication Greene et al. (2008) combined aspects of developmental and systems of beliefs models of personal epistemology while also drawing upon work in developmental psychology and phi­losophy to craft a model of epistemic and ontological cognition that we believe is a better reflection of the phenomena and more likely to result in a quantitative measure with strong psychometric properties and pre­dictive utility (see Greene et al. (2008) for a more complete descrip­tion of the model and its conceptual, philosophical, and methodological foundations). This model is intended to meet Hofer’s (2001) call for “an integration of ideas from multiple models: an identifiable set of dimensions or beliefs, organized as theories, progressing in reasonably predictable directions, activated in context, operating as epistemic cognition” (p. 377). We use the terms epistemic and ontological cog­nition rather than personal epistemology or epistemological beliefs

Using cognitive interviewing374

(Schommer, 1990) based upon two arguments. First, Kitchener (2002) has convincingly shown that beliefs about the nature of knowing are best described as epistemic, not epistemological. Second, Murphy and colleagues (Murphy, 2005; Murphy and Alexander, 2004) have demon­strated that beliefs about the nature of knowledge concern individuals’ ontology of academic domains, i.e., the categories and attributes within a domain and the relations among those categories (Pollock and Cruz, 1999; Slotta and Chi, 2006). Based upon these arguments, within this model, individuals’ cognition regarding the nature of knowledge (i.e., whether it is simple and certain or complex and dynamic) is termed ontological, whereas their cognition about the nature of knowing (i.e., justification) is termed epistemic, definitions that others have adopted as well (Hofer, 2008; Kitchener, 2002; Schraw and Olafson, 2008).

Philosophical epistemology “has traditionally focused on epistemic justification more than on knowledge” (Pollock and Cruz, 1999, p. 11). Indeed, as outlined in more detail by Murphy and colleagues (2007), most philosophical epistemologists define knowledge as justified true belief, with the specifics of the justification condition being the most contentious. The truth and belief conditions are relatively uncontro­versial in that few philosophers debate that knowledge requires belief and must be true (Plantinga, 1993; Pollock and Cruz, 1999; Williams, 2001). Yet in most conceptual models, justification for knowing is con­sidered but one aspect of personal epistemology, with equal or more attention paid to issues regarding the nature and sources of knowledge. To better align with the focus of philosophical epistemology, in our model of epistemic and ontological cognition issues of justification are the focus. Means of justification are measured using two separate dimensions. The justification by authority dimension captures how strongly individuals believe that knowledge claims can be sufficiently warranted using evidence from scientists, teachers, and other experts. The personal justification dimension measures how much faith people have in more idiosyncratic warrants, such as personal experience and logic.

We integrate multiple personal epistemology models by positing direct relations between dimensional and positional aspects of individuals’ epistemic and ontological cognition. We measure epistemic and onto­logical cognition along three separate but related dimensions, and posit that individuals’ beliefs along these dimensions will vary systematically, indicating their position within the model (see Table 12.1). We argue this position will be a stronger predictor of various outcomes of interest than any of the individual dimensions on their own. The three dimen­sions in the model are the two justification factors previously described

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 375

and a third capturing the degree to which the individual views knowl­edge as simple and certain. There are four positions in the model, based on the work of Chandler and colleagues (2002; Hallet et al., 2002) as well as that of Kuhn and colleagues (2000; Kuhn, 2005).

The first position, realism, conceptually resembles Kuhn and colleagues’ absolutism, and is characterized by strong belief in (1) knowledge as simple and certain; (2) justification by authority; and (3) personal means of justification. Individuals in this position believe that knowledge is objectively knowable, seeing warrants from either authority figures or the self as sufficient support for knowledge claims (i.e., strong belief in all three dimensions; see Table 12.1). As individu­als develop, they come to realize that the world is complex and dynamic (i.e., weak belief in simple and certain knowledge; see Table 12.1), lead­ing to a crisis: if people can legitimately disagree, then their justification for knowledge claims matters. Individuals resolve this crisis by moving into one of two possible positions. Dogmatists have strong belief in jus­tification by authority as a means of determining knowledge qua knowl­edge, and have weak belief in personal means of justification. On the other hand, skeptics justify their knowledge in a complex and dynamic world by relying solely on personal warrants, rejecting those of outside authorities.

Finally, some, but not all, individuals move past the dogmatist or skeptic position and into rationalism. Rationalists see the world as com­plex and dynamic, and have a strong need to evaluate the justifications underlying knowledge claims. However, they also recognize that dif­ferent types of justification may be more or less valid depending upon the circumstances. They have faith in their ability to evaluate warrants from authority figures as well as their own experience and cognitive processes, and acknowledge that in some circumstances each of these sources of justification may be more or less reliable. In terms of meas­urement, the model suggests that rationalists would have moderate agreement with items supporting justification by authority or personal justification, as opposed to dogmatists and skeptics who would strongly endorse one set of items and reject the other.

We also posit that epistemic and ontological cognition are not domain­general, in line with numerous other researchers’ claims (Buehl et al., 2002; Hallet et al., 2002). Rather than positing domain­specificity at the level of individual academic subjects, we propose that epistemic and ontological cognition varies according to whether the subject is well­ or ill­structured (Donald, 1990; Frederiksen, 1984; Stodolsky et al., 1991). Well­structured domains, such as math and the hard sciences, are gen­erally viewed as more objective with clearer rules of justification than

Tab

le 1

2.1.

Mod

el o

f epi

stem

ic a

nd o

ntol

ogic

al c

ogni

tion

Il

l­st

ruct

ure

d do

mai

ns

Wel

l­st

ruct

ure

d do

mai

ns

Edu

cati

onal

leve

lP

osit

ion

Bel

ief

in …

Pos

itio

nB

elie

f in

…

SC

JAP

J

SC

JAP

J

Ear

ly e

lem

enta

ry

sc

hool

Rea

lism

Str

ong

Str

ong

Str

ong

Rea

lism

Str

ong

Str

ong

Str

ong

Lat

e el

emen

tary

scho

ol t

o ea

rly

colle

ge

Dog

mat

ism

OR

sk

epti

cism

Wea

k W

eak

Str

ong

Wea

kW

eak

Str

ong

Rea

lism

Str

ong

Str

ong

Str

ong

Mid

dle

to la

te

co

llege

Rat

iona

lism

Wea

kM

oder

ate

Mod

erat

eD

ogm

atis

m

O

R

skep

tici

sm

Wea

kS

tron

g

Wea

kW

eak

S

tron

g

Pos

t­un

derg

radu

ate

educ

atio

nR

atio

nalis

m

Wea

k M

oder

ate

Mod

erat

e R

atio

nalis

m

Wea

k M

oder

ate

Mod

erat

e

SC

= s

impl

e an

d ce

rtai

n kn

owle

dge

dim

ensi

on; J

A =

just

ific

atio

n by

au

thor

ity

dim

ensi

on;

PJ

= p

erso

nal

just

ific

atio

n d

imen

sion

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 377

ill­structured domains, such as history and the social sciences. These domains are seen as more subjective in both content and rules of jus­tification. The differentiation of well­ versus ill­structured domains is important because the model posits that developmental change begins earlier for ill­structured domains than it does for well­ structured ones. As shown in Table 12.1, movement from a realist position is not posited to begin until late elementary school at the earliest for ill­ structured domains, but not until college for well­structured ones. In this way, the model aligns with findings in developmental psychology that elementary­ and secondary­school­aged students do understand the subjective aspects of certain types of knowledge (i.e., ill­structured domains), and can privilege certain justifications over others (see Hallet et al., 2002, and Flavell, 2004).

In sum, Greene and colleagues’ (2008) model of epistemic and onto­logical cognition is an important step toward Hofer’s (2001) call for an integrated model of personal epistemology that combines developmental and multidimensional aspects of other models by making specific claims regarding the relations between these aspects as well as how they change over time. The dimensions that characterize positions are informed by philosophical epistemology, in particular its focus upon justification. Further, the model allows for a degree of domain­specificity, at the level of well­ versus ill­structured domains. Finally, the model contains specific predictions regarding development, and to some degree runs contrary to other models of personal epistemology that claim develop­ment does not occur until college (e.g., Baxter Magolda, 2004; King and Kitchener, 2004; Perry, 1970).

Before we set out to test our model, we decided that any new quan­titative measure would need to be carefully piloted with elementary and secondary school students in order to gain a better understanding of these students’ epistemic and ontological cognition. We turned to cognitive interviewing as a means of investigating both the suitability of our instrument as well as the concepts that formed its foundation.

Cognitive interviewing and questionnaires

A powerful method for evaluating the suitability of items within ques­tionnaires is cognitive interviewing (Beatty and Willis, 2007; Karabenick et al., 2007; Willis, 2005). While definitions vary, in general cognitive interviewing involves administering questionnaires to a test sample of participants while also collecting verbal information about both the participants’ responses as well as their overall experience complet­ing the questionnaire. These verbal data can address issues such as

Using cognitive interviewing378

how participants constructed their answers, how they interpreted the items, what kinds of difficulties they had responding, if any, and any other information that can be useful in understanding the bases of the respondents’ answers (Beatty and Willis, 2007). We hoped that by using cognitive interviewing to identify variation in item­level interpre­tations we would reduce the measurement error of the questionnaire (Karabenick et al., 2007). Cognitive interviewing has also been used in the social sciences to gain insights into respondents’ experiences taking the questionnaire. This research can also shed light upon young peo­ple’s understanding of the constructs those questionnaires attempt to capture (see Richardson, 2006, for an example investigating students’ understanding of political discussion).

Karabenick and colleagues (2007) explicitly mentioned epistemo­logical beliefs as an area in need of cognitive interviewing research to “improve measurement validity and inform the meaning of theoretical constructs” (p. 139). We agree that cognitive interviewing has much to offer as means of gathering data regarding the validity and reliability of scores from survey­based measures of personal epistemology, as well as the more conceptual debates regarding construct definition, domain specificity, and development. According to Karabenick and colleagues, cognitive interviewing is also important when developing measures of abstract concepts to be used with younger respondents. Therefore, we felt it particularly appropriate to introduce this technique in a book focusing upon personal epistemology in preschool to grade twelve education.

The information from cognitive interviewing can be gathered either during or after questionnaire administration. Verbal protocols (Ericsson, 2006; Ericsson and Simon, 1993) are used most often to obtain information concurrently with item response. In these proto­cols, participants are asked to “say everything [they] are reading and thinking” as they respond to items. These verbalizations can surface conceptual understandings and misunderstandings, confusion regard­ing item wording, potential biases or confounds, and interest or bore­dom, among other things. Ericsson and Simon (1993) have provided evidence that the mere verbalization of the contents of working memory does not interfere with cognitive processes, but that directing partici­pants to explain why they are doing or thinking certain things can lead to interaction effects between the protocol and those cognitive proc­esses. Thus, most cognitive interviewing is done without requesting the participant to explain his or her thoughts, but merely verbalize them.

Cognitive interviewing often includes a semi­structured interview with retrospective probing after completion of the questionnaire to

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 379

capture information that is likely to be missed in concurrent verbal pro­tocols, due to the fact that participants are not asked to explain their reasoning while responding (Beatty and Willis, 2007). These retro­spective probes have the advantage of not interfering with participants’ responses to the questionnaire items, and can pertain to individual items, response scales, instructions, and overall impressions of the questionnaire that may not surface during a concurrent verbal proto­col. It should be noted that data gathered through verbal protocol can be analyzed in different ways depending upon the theoretical frame­work and goals of the investigation. In this study we used concurrent verbal protocol data collection within a cognitive interviewing design to understand participants’ conceptualization of the items, requiring thematic analysis akin to qualitative techniques. Although not done in this study, verbal protocols can also be used as a trace methodology to gather data as students learn. In the latter use of verbal protocols, the data can be analyzed using coding schemes to identify relevant cogni­tive, metacognitive, and motivational processes used during learning and problem­solving (e.g., Greene and Azevedo, 2007).

Thus, cognitive interviewing can address both the adequacy of the questionnaire itself as well as deeper conceptual understandings by using concurrent verbal protocols and semi­structured interview­ing with retrospective probes, as the two “actually fit together very naturally” (Willis, 2005, p. 57). Other methodological aspects of cognitive interviewing are less clear, such as specifics regarding nec­essary sample size and diversity, types and number of probes, and determining the validity of inferences from participants’ verbal data (Beatty and Willis, 2007). Sample sizes for cognitive interviewing are generally small, ranging from five to fifteen individuals, and include only enough participants to identify problems with the questionnaire that may be prevalent in the larger population, as well as informa­tion regarding those specific participants’ conceptual understandings (Beatty and Willis, 2007). The idea of saturation (Glaser and Strauss, 1967) is one criterion for determining whether the sample size is suf­ficient. When a researcher determines that the relevant information has been captured, and further interviews have a low probability of bringing about new insights, the data are considered saturated and the interviewing ended. Criteria for sample diversity include whether participants vary on relevant characteristics, ensuring a broad range of responses.

The probes used can be written prior to interviewing, or can emerge from the researchers’ observation of the participant during question­naire completion (Beatty and Willis, 2007). It is advisable to have probes

Using cognitive interviewing380

that tap the pertinent interpretive and conceptual understandings of participants while also not being too leading or bias­inducing. Willis (2005) recommends that emergent probing be done only by those who are well­versed in the ideas underlying the questionnaire, as opposed to paid questionnaire administrators, who may lack the needed theoretical background to develop probes based upon observation. Ideal probes are open­ended, allowing participants the freedom to answer as they wish while also requiring more than a Yes/No response.

Finally, the concurrent verbal and semi­structured interview data must be analyzed, and inferences based upon these analyses must be scrutinized for validity. Unfortunately, there are few established meth­ods for analyzing cognitive interview data (Willis, 2005). One method involves combing verbal data for common themes across participants, using qualitative analysis coding procedures (see Creswell, 2003). How­ever, in some cases only one participant may verbalize a misconception regarding an item, and the researcher must decide whether that is suf­ficient evidence to warrant changes to the questionnaire. Ultimately, Beatty and Willis (2007) recommend relying less on the number of par­ticipants who express a concern or misunderstanding, and more upon whether a logical argument can be made that the participant’s concern is legitimate and not due to an idiosyncratic issue that is unlikely to be typical of the larger population.

Research purpose and questions

This study involves additional analyses of a subset of the data collected in support of Greene’s (2007) work on epistemic and ontological cogni­tion. In this study, cognitive interviewing was used for two purposes. First, we sought to gather information regarding the questionnaire itself, the quality of the items, and the response scale. Second, we wished to use the participants’ concurrent verbal and semi­structured interview data to better understand elementary and secondary school students’ actual epistemic and ontological cognition. We sought data regarding the actual constructs under study, the range of response within these populations, the sophistication of these students’ thinking regarding epistemic and ontological issues, and their views regarding various means of justification.

Our research questions were:

(1) How do elementary and secondary school students experience and interpret the various aspects of our data collection and instrument (e.g., the concurrent verbal protocol, items, response scale, domains covered)?

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 381

(2) Is the response scale clear to elementary and secondary school students?

(3) Do elementary and secondary school students think differently about well­ and ill­structured academic domains in terms of our three dimensions of epistemic and ontological cognition?

(4) What is the nature of elementary and secondary school students’ epistemic and ontological cognition?

Method

Participants

Four secondary school students and three elementary school students were successfully recruited for this study using purposeful sampling during the spring and summer of 2006. The secondary school students included two attending a public high school in suburban Maryland, a student attending a private school in Washington, DC, and a student attending a military school in Virginia. The elementary school students all attended schools in suburban Maryland, two of which were public and one private. Demographic information regarding the participants can be found in Table 12.2. The sample was predominantly male, which in some research designs would lead to a concern regarding the general­izability of the findings, but given the goals of cognitive interviewing is less of a concern here (Beatty and Willis, 2007). However, this skew is a potential limitation in terms of identifying possible concerns that are more prevalent in females.

Instrument

A pilot instrument, the epistemic and ontological cognition question­naire (EOCQ), was created to measure the three dimensions outlined in our model of epistemic and ontological cognition (see Table 12.1; Greene et al., 2008). All of the items created were Likert­type (Ueber­sax, 2006) with response scales ranging from one to seven, with one being labeled as “completely disagree,” four being labeled as “neither disagree nor agree,” and seven labeled as “completely agree.” We strove to phrase items strongly to ensure that participants’ degree of agree­ment or disagreement with the item would be captured in their choice of response option, and not confounded by interpretations of the lan­guage of the item (DeVellis, 2003).

Content areas. In our model we predict that epistemic and ontological cognition will vary between well­ and ill­structured domains (Greene

Using cognitive interviewing382

et al., 2008). To test this prediction, we sought out academic areas that matched Hallet and colleagues’ (2002) assertion that “the long­stand­ing, if somewhat controversial, distinction between the social and natu­ral sciences … is … another instantiation of the difference between” (p. 293) well­ and ill­structured domains. We decided upon four aca­demic areas, physics and math for well­structured domains, and his­tory and political science for ill­structured domains. We had hoped that participants would be familiar with each, thus facilitating a test of hypothesized between group differences (i.e., physics and math scores compared to history and political science scores) as well as within group similarities (i.e., physics and math scores were predicted to be similar, as were history and political science scores). We wrote each item such that references to one domain (i.e., physics, math, history, political sci­ence) could be substituted with each of the others, making items paral­lel across those domains. For example, four versions of this item were used, with each differing only in terms of the domain listed: “In [phys­ics, math, history, political science], the truth means different things to different people.”

Simple and certain knowledge dimension. The simple and certain knowl­edge dimension was based upon the work of Schommer and Schraw and colleagues; thus it seemed reasonable to examine their instru­ments for potential items. The two measures used by these research­ers are the epistemological questionnaire (EQ; Schommer, 1990) and the epistemic belief inventory (EBI; Schraw et al., 2002). Numerous studies have provided information regarding the functioning of items from these measures, supplying construct validity evidence such as factor loadings (Schommer­Aikins et al., 2000; Schraw et al., 2002). As we selected items, we also considered content validity. We endeav­ored to include items that captured the breadth and depth of the phe­nomena, for example including items concerning the factual aspects

Table 12.2. Participant demographic information

Participant number Education level Sex Age

01 Secondary school Male 1802 Secondary school Female 1703 Secondary school Male 1704 Elementary school Male 1005 Secondary school Male 1606 Elementary school Male 1007 Elementary school Male 10

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 383

of academic areas, issues of truth, and items that asked participants to assess the changing nature of knowledge. This review led to the inclu­sion of four items from the EBI, each adapted from a domain­general wording to phrasing appropriate for well­ and ill­structured domains. The original items and the versions used in the EOCQ can be found in Table 12.3.

The “simple” aspect of this dimension was measured using items referring to “truth” or “facts” and whether those things could differ across people. Items concerning whether knowledge would always be true in the future and whether facts could change were included to capture the “certain” aspect of this dimension. An example item for this dimension was “In math, the facts do not change.” Four of the items were written such that strong agreement with the item indicated a more naïve stance, i.e., believing knowledge was simple and certain. Two additional items measuring this dimension were reverse­coded, i.e., they were written such that strong agreement would indicate more sophisticated beliefs regarding the nature of knowledge.

Justification by authority. Five items were created to measure the jus­tification by authority dimension, with two reverse­coded. These items

Table 12.3. EBI wording and EOCQ wording

EBI wordinga EOCQ wording

Well­structured domain Ill­structured domain

If two people are arguing about something, at least one of them must be wrong.

If two [scientists, mathematicians] disagree about some part of [physics, math] one of them must be wrong.

If two [historians, political scientists] disagree about some part of [history, political science] one of them must be wrong.

What is true today will be true tomorrow.

In [physics, math], what is true today will be true tomorrow.

In [history, political science], what is true today will be true tomorrow.

Sometimes there are no right answers to life’s big problems.

There are some things in [physics, math] that we will never understand.

There are some things in [history, political science] that we will never understand.

Science is easy to understand because it contains so many facts.

To do well in [physics, math] class, the main thing you need to do is memorize facts.

To do well in [history, political science] class, the main thing you need to do is memorize facts.

a As listed in Schraw et al. (2002)

Using cognitive interviewing384

were intended to measure the degree to which participants put their faith in authority figures and the degree to which they accepted, without question, the claims made by teachers, domain experts, and in textbooks. In our view, textbooks were important to include as a more non­human manifestation of authority. An example of item measuring this dimen­sion was “If a mathematician says something is a fact, I believe it.”

Personal justification. For us, the most difficult items to draft were those intended to measure participants’ belief in personal means of jus­tification. It was challenging to construct items that assessed partici­pants’ faith in their own means of justification, to the exclusion of all others. Many of the items we crafted initially sounded forced or lead­ing, and were discarded. We did manage to create three items focused upon individualistic means of justification and participants’ acceptance of them as legitimate warrants for knowledge claims (e.g., “In history, you just have to decide what you believe and what you do not”). One of these items was reverse­coded, and measured whether individuals believed knowledge could be justified qua knowledge at all.

Measure length. There were a total of fourteen items across all three dimensions, and with parallel items for each of the four academic domains this resulted fifty­six items. In addition, items regarding par­ticipants’ beliefs regarding aesthetics were included (but not reported here, since those items were dropped from the instrument). Finally, there were five demographic items, bringing the total number of items on this pilot version of the EOCQ to seventy­three.

Procedure

The first author did all of the cognitive interviewing, which involved concurrent verbal protocols and semi­structured interviewing with retrospective and emergent probes. Before conducting the cognitive interview, participants read and signed an assent form, and a parent or guardian read and signed a consent form. At the beginning of the session, the interviewer explained that the purpose of the study was to examine the participant’s ideas regarding knowledge and knowing, as well as to learn more about how the participant interpreted the items on the questionnaire. The interviewer stressed that there were no “right” answers to the items, and that if an item was not clear that it was not an indication of any deficit on the part of the participant, but rather a problem with the item itself.

After answering any initial questions, the interviewer reviewed the concurrent verbal protocol (Ericsson, 2006; Ericsson and Simon,

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 385

1993) with the participant. Participants were asked to read aloud, and to say everything they were thinking while taking the question­naire. The interviewer demonstrated the concurrent verbal protocol, and emphasized that all thoughts should be verbalized. The partici­pants were told that if they were silent for more than three seconds, the interviewer would prompt them by saying, “Can you say what you are thinking?” After answering any questions regarding the concurrent verbal protocol, the interviewer asked participants to put on the micro­phone, turned on the audiotape recorder, and requested participants to begin.

The questionnaire itself was administered in paper­and­pencil format, and participants were given as much time as they needed to fill it out. The interviewer observed the participants as they filled out the ques­tionnaire, taking field notes and crafting emergent probes. After com­pleting the questionnaire, the interviewer initiated the semi­structured interview, consisting of both prewritten (retrospective) and emergent probes (Beatty and Willis, 2007). Examples of the prewritten probes are shown in Table 12.4, and parallel those recommended by Karaben­ick and colleagues (2007) when interviewing children. Emergent probes were created from the interviewer’s observations of the participant while completing the questionnaire, as well as the participant’s concurrent verbal protocol data. For example, if a participant verbalized a thought about the changing nature of knowledge in history, the interviewer created an emergent probe to explore this belief. The semi­structured interview was also tape­recorded. After this interview, the interviewer thanked the students for their participation, stopped the audiotape, and answered any questions.

Data analysis

Transcription of audiotapes. The interviewer transcribed each audio­tape, which included both the concurrent verbal protocol and semi­ structured interview data. This produced a corpus of 106 double­spaced pages of text (M = 15.1) with 3,469 lines (M = 495.6) and 26,219 words (M = 3745.6). These verbal data were also reorganized by item

Table 12.4. Examples of prewritten semi-structured interview probes

What are your overall impressions of the questionnaire?Were there any items that you did not understand?Were there any response options you wished had been there?What did you think the questionnaire was trying to find out about you?

Using cognitive interviewing386

to facilitate analysis. In addition, the field notes were consulted during the analyses.

Analysis procedures. As stated previously, data analysis in cognitive interview studies is not standardized and often guided by an eclectic set of principles derived from previous research (Beatty and Willis, 2007; Willis, 2005). The purpose of our data analyses was to understand not only the participants’ experience completing the questionnaire itself but also to gain a better perspective on their epistemic and ontological cognition in general through their responses to items as well as through the semi­structured interviews that followed. The concurrent verbal protocol and semi­structured interview data were examined for text segments that could be coded and combined into themes (Creswell, 2003). Codes can either be determined a priori (called the system­atic approach; Strauss and Corbin, 1998) or generated from the data inductively (called the emergent approach; Glaser, 1992). In our case, we entered into the data analysis looking for evidence of epistemic and ontological cognition, but did not create any specific codes or themes a priori. Using this emergent approach, our codes were derived from the data, first by examining each transcript individually and then look­ing across transcripts for commonalities and differences. We then col­lapsed those codes into a smaller number of themes that captured the important findings in the data.

Efforts to establish the validity and credibility of the data and associ-ated inferences. Validity in cognitive interviewing and qualitative data analysis concerns the credibility or trustworthiness of both the data gathered and the inferences made from such data (Beatty and Willis, 2007; Creswell, 2003; Patton, 1990). Qualitative data collection is con­sidered more valid when the researcher takes the time to establish rap­port with participants, which the interviewer did prior to the cognitive interviewing. In addition, we used audio recording to ensure that the participants’ actual words were captured and analyzed, rather than interviewer notes alone. Use of audio recording also allowed the first author to focus upon analyzing participants’ concurrent verbal proto­cols in real­time, facilitating the construction of field notes and relevant emergent probes to be used in the semi­structured interviews.

Two separate triangulation techniques (Patton, 1990) were used to bolster validity. In this study, triangulation of data sources was achieved by verifying the presence or absence of themes across multiple partici­pants. Likewise, certain themes could be verified by examining the first author’s field notes taken during observation (Creswell, 2003). Mul­tiple analysts were also used to triangulate the themes and findings. When more than one analyst uncovered similar themes, this was taken

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 387

as support for the inference. The first three authors independently examined the findings for coherence with both theory and the data. In addition, the fourth author performed an external audit, creating her own codes and themes for all of the transcripts. The themes generated by the first author were corroborated or extended by the fourth author’s work, providing further validity evidence.

Results

We divide our results into two sections. Our first two research ques­tions concerned methodological and measurement issues regarding the concurrent verbal protocol, the response scale, and participants’ dif­ficulties with certain academic domains. We provide several examples that illustrate why cognitive interviewing is important when developing items. In general, methodological and measurement findings required little inference and were mostly based upon observation and the nature of the participants’ responses. While findings regarding the phrasing of various specific items are discussed when appropriate, the second section includes data and inferences more related to the study of epis­temic and ontological cognition in general, and addressed our third and fourth research questions. These findings required qualitative analy­sis, including inferring general themes from multiple codes. We review five main themes: (1) domain­level differences in ontological cognition; (2) ontological sophistication in history; (3) continua of secondary stu­dents’ ontology and authority views; (4) secondary students’ varying faith in types of authorities; and (5) secondary students’ differing upon the nature of the terms “true” and “truth.” Each of these themes is further developed, below. For each of the themes, we felt the data were saturated.

Research questions 1 and 2: issues related to methodology and measurement in this study

Concurrent verbal protocol. Given that young children may struggle with concurrent verbal protocols (Beatty and Willis, 2007), it was not sur­prising that each of the elementary school students had an enormous amount of difficulty completing the questionnaire while verbalizing. Repeated prompts for students to verbalize were followed by silence or statement of the response number, rather than the participants’ thought processes. The elementary school students’ inability to verbalize their thinking required the first author to take careful field notes while observ­ing them complete the questionnaire. These field notes then informed the emergent probes used in the semi­structured interviews (e.g., “Were

Using cognitive interviewing388

there any questions in particular that stand out in your mind that you think you didn’t understand?”), and in general resulted in longer inter­views with the elementary school students as compared to the secondary school students. Secondary school students, on the other hand, did not display any problems with the concurrent verbal protocol.

Response scale issues. Each participant was asked whether the seven­point Likert response scale was interpretable and sufficient, and all agreed on both counts. The interviewer did not observe any partici­pants struggling with the number of response options. However, the interviewer observed numerous times when participants said they were unclear about what the item was asking, followed by them selecting “neither disagree nor agree.” Clearly there is a difference between not comprehending an item and feeling neutral in terms of its endorsement. Unfortunately, this kind of problematic behavior is often found when there are an odd number of response options (DeVellis, 2003). The best way to address this problem is to write items that are clear for all par­ticipants. But, short of this somewhat unrealistic ideal, another option would be to use scales with an even number of response options, and no true neutral (DeVellis, 2003).

Lack of familiarity with physics and political science. We included four versions of each item, addressing two well­structured and two ill­ structured domains, in the hopes of being able to compare participants’ views within these two classifications. Unfortunately, it was clear that the elementary school students had no familiarity with physics or pol­itical science:

p 04: Some of the questions I didn’t understand – the political science.j g : Okay, good, um what, what didn’t you understand or why were they why

were they hard for you?p 04: Um well, I haven’t done political science in school yet.j g : So you haven’t done it so it was hard, was it hard trying to guess?p 04: Yeah.j g : Okay, what about physics?p 04: I haven’t done physics in school either, but I can have an idea of what it’s

like.j g : Mm­hm okay, and what do you think it’s like?p 04: Um … it’s like I think it’s like learning a life, like how you’re live your

life, I’m not sure.

One of the secondary school students also struggled with these domains:

j g : Okay, so first of all what were your overall impressions of filling this out?p 05: I don’t know, I have no idea what political science … and I haven’t taken

physics so, I have no idea what, I didn’t really know what to put for those.

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 389

To some extent these findings are not surprising, particularly given that many students do not take physics or political science in elementary or secondary school. Clearly, these findings are informative in terms of our first research question. In a more general sense, these findings highlight a difficult issue regarding the measurement of epistemic and ontological cognition.

It can be argued that homogeneous samples, such as those includ­ing only college students, are likely to result in restriction of range issues. For example, in our model, only elementary school students are expected to have strong agreement with items measuring simple and certain knowledge in history. Samples that exclude these students will likely have a restricted range of response (i.e., mostly weak to strong disagreement). Restriction of range can attenuate the results of factor analyses used to test the validity of scores from quantitative measures of epistemic and ontological cognition (Gorsuch, 1983). Therefore, it is imperative that designers of questionnaires use academic domains that are familiar to elementary through graduate school students, so that necessary variance in the range of response can be captured. The further questionnaire designers get from the traditional academic sub­jects like math, history, and English, the more likely it is that some subset of their sample, particularly the younger students, will not be familiar with the domain, potentially introducing serious error into the measurement.

Other evidence supporting the importance of using cognitive interviewing when developing items. On the EOCQ, ontological cognition is captured using items assessing participants’ degree of belief that the nature of knowledge is simple and certain. One set of parallel items (“To do well in [math, history] class, the main thing you need to do is memorize facts”) was a variation of an item from the EBI (see Table 12.3). In regards to math, two of the secondary school students disagreed with this item, and one elementary school student responded neither disa­gree nor agree. These responses could be interpreted as indicative of a more sophisticated belief regarding the nature of knowledge in math, a finding contrary to the predictions of most models of personal epis­temology, including ours. However, the concurrent verbal protocol data revealed that these students were reacting specifically to the focus upon facts: “you only have to remember formulas, you don’t remember facts” (P02). This is a good example of why piloting items is important. The concurrent verbal protocol data and responses to emergent probes confirmed that despite some quantitative responses suggesting other­wise, no student displayed a sophisticated belief regarding the nature of knowledge in math.

Using cognitive interviewing390

Another item that was modified from the EBI (“If two historians dis­agree about some part of history one of them must be wrong”; see Table 12.3) also illustrated the value of piloting items with cognitive inter­viewing. Secondary school students’ quantitative responses were con­sistent with our conceptual model, but the concurrent verbal protocol data revealed a misinterpretation. The items were designed to measure whether participants believed that history was complex, but numerous participants explained that they disagreed with the item because both experts could be wrong. We had sought to measure ontological cogni­tion, but participants’ responses had more to do with semantics. The original EBI wording may be less likely to elicit this problem.

Research questions 3 and 4: findings informative of epistemic and ontological cognition in general

Our third and fourth research questions concerned more general issues including students’ beliefs regarding the nature of knowledge and knowing. We wished to test specific claims in our model (i.e., degree of domain­specificity) as well as to learn more about epistemic and onto­logical cognition in general.

Ontological domain-specificity and sophistication. While we could not completely test our hypothesis regarding well­ versus ill­structured domain differences in epistemic and ontological cognition, we did learn a great deal about how students’ beliefs differed across math and his­tory. For example, in terms of ontological cognition, the elementary and secondary school students saw math knowledge as objective and comprised of facts or static knowledge. We coded many of the par­ticipants’ verbalizations as “math just facts” and “math success equals memorizing formulas.” Likewise, one elementary and one secondary school student saw history as objective, with many of their verbaliza­tions coded as “history success equals memorizing facts,” “history claims are right or wrong,” and “history facts do not change.” On the other hand, the other two elementary school students showed a limited understanding of history’s subjectivity, recognizing the role of opin­ions or interpretations of historical events. The following two dialogues were instructive:

j g : Okay, great. Um … let’s see … so…for in history the truth means different things to different people you put a five, why did you put that?

p 04: Because um … with the civil war, um, some, all of the Americans think that um…we uh, we were like, yeah, we were like really good.

j g : Mm­mm.p 04: To, yeah, to … to do that, to start a war because it wasn’t fair.j g : Mm­mm.

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 391

p 04: But some, some people in England probably thought, maybe the oppos­ite of that.

j g : Oh, okay, oh the American Revolution?p 04: Yeah, American Revolution.j g : Right okay.p 04: That’s it.j g : So, um, so if in America people thought they were doing the right thing

but in England people may have thought they were not doing the right thing?

p 04: Yeah.j g : Okay, and um, is one of those groups more right or more wrong?p 04: Um … no.j g : No.p 04: No, cause…also in the Civil War too, because slaves, the slaves, some

people thought it was good to have slaves but some people thought it wasn’t right.

j g : Okay, and there’s not kind of a right or wrong answer?p 04: Yeah.j g : How much of history is opinion and how much of it is fact?p 06: Most of it, like … usually if something has been accomplished, the fact

would be that they have accomplished it.j g : Mm.p 06: If something’s been lost, like everywhere I go I see some picture of

someone putting up the American flag at the end of the Second World War that is a fact that we won.

j g : Mm­hm.p 06: But I’m not, the opinion would be who should have won, or like should

the war should have been started, or if anything should have been started about anything.

Both of these verbalizations were coded as “history composed of facts and interpretations.” Three of the secondary school students had what we considered a sophisticated view of knowledge in history, as evidenced by this quote: “A lot of histories is noticing patterns between stuff and being able to predict patterns” (P01). In response to the question “To do well in history class, the main thing you need to do is memorize facts” another secondary school participant said, “the main thing in history is to interpret the facts and understand why they happened” (P02). Given that these participants viewed math as objective and history as subjec­tive, these findings are supportive of domain­specificity in ontological cognition. In general, the codes in this area were subsumed under the themes “domain­level differences in ontological cognition” and “onto­logical sophistication in history.” Further, the finding that two of the three elementary and three of the four secondary school students had an ontology for history that included much more than a compendium of facts and dates contradicts the expectations of naivety found in many

Using cognitive interviewing392

personal epistemology models, and falls in line with work from develop­mental psychology and our conceptual model.

Continua of secondary students’ views of ontology and authority. The sec­ondary school students were familiar with all four domains, and thus able to respond to the physics and political science items, displaying a nuanced set of viewpoints. We interpreted their verbal data as indica­tive of a set of continua (see Figure 12.1). Secondary students often verbalized the belief that math was composed of facts, but that to some degree in physics, and even more so history, knowledge included both facts and interpretations. Political science was seen as almost entirely composed of opinions, which may reveal a naivety regarding the con­tent of this domain.

From the coded verbalizations we inferred the general theme “sec­ondary students beliefs’ about domains’ degree of objectivity exist on a continuum.” We think the following response to an emergent probe does a good job illustrating this continuum:

Well, just math and physics, I think it’s pretty much similar, like, those are more like facts that have been proved, like 99 percent … and history’s more like, well, it’s the past but you can put your own, uh, definition to it, some­times. Like political science is completely different because it’s like well … you are what you are and you think what you think. (P09)

Likewise, we saw a theme related to knowledge in these domains being viewed on a continuum from certain and static to tentative and chang­ing. For example, while most secondary school students saw math knowledge as unchanging, history claims were seen as subject to revi­sion: “they’re always finding out new facts about what may have led to this … something they’ve discovered is wrong or, it’s a lot of unearthed things that they don’t know about” (P01). Finally, these secondary stu­dents’ responses indicated relatively strong faith in the expertise and relative infallibility of math teachers, and more skepticism or doubt for teachers in other domains. Codes here included “I don’t trust history teachers,” “physics teachers can make mistakes,” and “math teachers have the right answer.” We found it particularly interesting that these students’ views of their teachers seemed to vary in line with their per­ceptions of the nature of knowledge in these domains, and inferred a theme identifying a continuum of faith in authority figures across the four domains.

Epistemic cognition: secondary students’ varying faith in types of authori-ties. In the EOCQ, epistemic cognition was measured in two ways. The first way involved items assessing participants’ faith in various forms of authority as justification for knowing. For the most part, participants’

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 393

quantitative responses to the justification by authority items aligned with the predictions of our model. However, the concurrent verbal pro­tocol data showed that secondary school participants held a much more nuanced understanding of authority figures than generally assumed in models of personal epistemology. Participant five, a secondary school student, strongly agreed with items assessing faith in authority figures such as teachers and books. However, several other secondary school participants completely disagreed with statements like “If a [mathe­matician, physicist] says something is a fact, I believe it,” verbalizing conditions such as “they have to show proof” (P02) and “just because he’s a physicist doesn’t mean whatever he says is true” (P03). These participants felt similarly regarding historians: “There must be proof, they just can’t say something and believe it” (P01). In another example, one elementary school student saw experts as fallible:

j g : So let’s see so, on then number 14 “if mathematician says something is a fact I believe it” and you also said neither disagree nor agree, why did you say that?

p 07: Um, cause sometimes people can be wrong, sometimes, too.j g : Okay.p 07: They might forget something, [ … ] or get it wrong.

Another elementary school student had a similar response, saying that “mathematicians aren’t necessarily a calculator” (P04). Interest­ingly, this student viewed teachers as less authoritative than experts or computers:

j g : Okay so this one “If my friend and I disagree about something we learned in physics class, we just have to ask the teacher to tell us who is right” and you put more disagree than agree, can you tell me why you put that?

Math Physics History Political science

Facts Interpretations and facts Opinions

Certain knowledge Uncertain knowledge

Teachers arealways correct

Teachers can be wrong

Figure 12.1: Relative location of academic domains on secondary students’ belief continua

Using cognitive interviewing394

p 04: Because you can maybe like look it up on the computer and find out the true answer than just going back to your teacher and asking or some­thing.

j g : Okay, so you think that the computer is a better, the computer um is a safer way of finding the truth than asking your teacher?

p 04: Yeah.j g : Okay why do you think that?p 04: Um, because your teacher may not know exactly what is true but um the

computer has like experts on it.j g : Okay, so, you don’t, um, do you think teachers can be experts?p 04: They could be.j g : But not necessarily?p 04: Yeah.

A secondary school student responding to the same question said, “I don’t always trust the teacher is a very good teacher.” Math and history textbooks were also seen as more authoritative than individual teachers because they have “things that happened put in by many peo­ple and if they are in a textbook then it’s because it’s something close to truth … so I don’t believe all the things [my teacher] says, I refer to my textbook if I don’t know something” (P02). We believe these findings illustrate an important theme regarding secondary students’ view of authority in all academic areas except math: there is evidence of a hier­archy, ranging from teachers being least compelling to textbooks and computer sites being the most compelling, mainly due to the perception that the latter combine the views of multiple sources.

Epistemic cognition: secondary students’ differing interpretations of the word “true.” We also captured participants’ epistemic cognition concerning more personal forms of justification. The items written for this dimen­sion were intended to assess participants’ faith in their own personal experience and rationality as means of justification. As a whole, they were the most poorly performing items on the questionnaire. The ele­mentary school students, who were expected to agree with these types of justifications, had a great deal of trouble understanding the items, with one unable to even provide a quantitative response or an explana­tion as to why he struggled with the items. The other elementary school students’ responses were contrary to our model­based predictions, and follow up probes suggested that the responses were due to confusion based upon the way the items were written, and not due to any unex­pected sophistication. A majority of the secondary school participants’ responses were also contrary to the model, with the math domain items very problematic and in need of revision. Again, we determined that these unexpected responses were due to the participants interpret­ing the items differently than we intended. In particular, participants

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 395

inferred that the items were referring to opinions, and not knowledge claims, primarily because the items included the words “believe” and “true.”

These findings are not surprising. Alexander and colleagues found that secondary school students had somewhat different implicit defini­tions of words like “knowledge” and “beliefs” (Alexander and Dochy, 1995; Alexander et al., 1998). While most students agreed the con­cepts had some overlap, knowledge was seen as more objective whereas beliefs were viewed as more subjective. Knowledge was associated with words like “know” and “facts,” whereas students’ definitions of “beliefs” most often included words such as “true” and “values.” From our study, it was clear that this distinction led to problems when assess­ing the reliability of personal means of justification (i.e., “In history, you just have to decide what you believe and what you do not”) but was less influential when the item concerned participants’ assessment of an authority figure (i.e., “If a mathematician says something is a fact, I believe it”).

In our study, words that Alexander and colleagues found were asso­ciated with beliefs, in particular the word “true,” were interpreted in reference to beliefs when participants spoke of their own justification processes, but were associated more with knowledge when used in ref­erence to authority figures. Future research should investigate whether students interpret the word “true” differently depending upon whether it is used in reference to the self or authority figures.

Given our findings, in our revisions to the personal justification items, we removed terms such as “true” and “truth” to avoid confusion between knowledge claims and issues of values or morals. The revised version of the EOCQ has since been used in a large­scale administra­tion including over 700 participants, with quantitative analyses showing support for many aspects of the underlying conceptual model (Greene and Azevedo, 2008). Further item revision may be warranted given the quantitative analyses.

Implications

Our findings demonstrate the value of cognitive interviewing in terms of understanding the phenomena of epistemic and ontological cogni­tion as well as the means by which they are measured. We discuss how specific findings inform conceptual models of personal epistemology literature and measurement of epistemic and ontological cognition. We also review implications for elementary and high school students’ education.

Using cognitive interviewing396

How findings inform conceptual models of personal epistemology

In terms of the personal epistemology literature in general, the most important finding is evidence that some secondary school students do display rather sophisticated beliefs regarding the nature of knowledge and knowing in history. Three of our secondary school students provided evidence that epistemic and ontological cognitive development, moving away from dualism, can begin earlier than the college years and without the stimulation of college­level classes. It would appear that college is but one of many of snakes promising knowledge, to use Perry’s (1999) analogy, and students are banished from their simple and certain Eden much earlier than their freshmen year. The causes of this precollege development are not clear. It may be that students are developing earlier, or that the secondary school experience is more “college­like” than when Perry gathered his data. Certainly the relatively recent focus upon con­structivism in teacher education may be affecting how students are asked to approach knowledge in domains such as history (VanSledright and Limón, 2006). Another explanation may be that the Internet has caused students to adopt more sophisticated beliefs regarding knowledge.

Conceptual models of personal epistemology should be expanded to allow for development prior to college, incorporating our findings regarding elementary and secondary school students’ understanding of the role of opinion in history. While this might simply mean allow­ing more domain­specific differences regarding personal epistemol­ogy, the implications of such a change are many. Hofer (2004b) has hypothesized that more naïve beliefs regarding the nature of knowledge may be associated with less sophisticated strategy use during learning. Does the level of historical understanding (VanSledright, 2004) that some elementary students displayed make them more likely to invoke sophisticated strategies than students lacking this understanding? Also, are there multiple latent continua regarding the nature of knowledge, perhaps representing a person’s degree of belief regarding the pre­dominance of facts, opinions, and interpretations in a given academic domain? Finally, our data provide further evidence that individuals’ beliefs about the nature of knowledge and knowing are not domain general and should not be modeled as such (Alexander, 2006; Buehl et al., 2002; Hofer, 2006; Muis et al., 2006).

How findings inform measurement

All of the findings discussed previously have implications for improv­ing the measurement of epistemic and ontological cognition, including

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 397

participants’ varied interpretations of words like “believe” and “true,” their domain­specific responses, their tendency to place more faith in sources of knowledge with multiple authors (i.e., textbooks), their beliefs regarding the differential fallibility of teachers in different domains, and their recognition of the role of opinion and interpretation in history. In general, these findings suggest that many current quan­titative measures of personal epistemology may need to be revised to take into account these issues. For example, domain­specificity can be addressed by following Kuhn and colleagues’ (2000) work and making parallel assessments for multiple domains. Also, it may be necessary to use Likert response scales with an even number of options, to prevent the true neutral response being used as a proxy for “I don’t know” (potentially an epistemological stance on its own that is certainly not half way between agreement and disagreement).

Likewise, our findings raise the possibility of an interaction between academic domain and the justification by authority factor. It may be that secondary school students have relatively little faith in author­ity figures in ill­structured domains, not because they have adopted a skeptical or multiplistic belief regarding the nature of knowing but rather because they have had experiences with teachers in these domains who admitted fallibility. This suggests the need for a more nuanced set of factors measuring faith in authority figures, taking into account perceptions of expertise as well as issues of fallibility. This also means it is important to measure students’ beliefs regarding a variety of sources of authority besides teachers. A realist or dogmatist position is typified by faith in authority figures, but we have evidence suggesting that secondary school students moderate this faith based upon the perceived number of authorities contributing to the source. If textbooks and computer websites are seen as reliable amalgams of multiple experts, this is another indicator of potentially naïve thinking that needs to be assessed. This information may be missed if items designed to measure justification by authority only include individual persons as sources.

Through the concurrent verbalizations we found a misinterpretation that calls into question many of the items used in the literature. Sec­ondary students seemed to interpret words like “believe” and “true” as referring to opinions when used in reference to themselves, but as refer­ring to knowledge or facts when used in reference to authority figures. This is intriguing, and may expand upon Alexander and colleagues’ (1998) findings regarding a substantial degree of overlap between students’ conceptions of knowledge and beliefs. In terms of measure­ment, our findings suggest that items containing these words must be

Using cognitive interviewing398

carefully vetted, particularly in terms of whether they are intended to assess faith in authority versus the self.

In general, our study illustrates the value of using cognitive inter­viewing to pilot measures. It is the case that one rarely sees publications referring to any kind of item testing besides factor analyses and internal reliability measures, which are extremely important but insufficient. As Richardson (2006) noted, these pilot studies may be conducted but simply not reported. This study illustrated that at times participants respond to a quantitative scale in a way that aligns with theory, but does not validly capture the nature of their beliefs. Epistemic and ontologi­cal cognition are especially abstract and difficult to measure, so there should be a razor­sharp focus on developing unambiguous items, and cognitive interviewing can be an effective means of accomplishing this. Our measure continues to be tested and refined in this manner, using both quantitative and qualitative analysis techniques.

Educational implications

Educational implications of our model. A model of epistemic and ontologi­cal cognition that accurately describes students’ beliefs about knowledge and knowing should be generative, leading to more informed pedagogy and learning. We believe our model has a number of important implica­tions. First, our description of multiple means of justification suggests that teachers need to understand not just what students think they know, but why they think they know it (Hofer, 2001). It may be that interven­tions designed to foster conceptual change must vary depending upon whether the students’ justification for their misconception is based in authority or personal experience (Chinn and Brewer, 1993; Dole and Sinatra, 1998; Murphy, 2007). Likewise, attribute­treatment interac­tions may exist where students who only acknowledge justification from authority figures (dogmatists) may do poorly in problem­based learn­ing environments that ask students to seek their own answers (Hmelo­Silver et al., 2007).

A second implication of our model is that assessment of individual beliefs about the nature of knowledge and knowing are less effective at predicting academic outcomes than using those assessments in com­bination to determine a student’s overall position (see Table 12.1). For example, determining that a student has strong belief in personal means of justification has limited utility under our model. Such a student could be a realist or a skeptic. Determining which position the student is in requires additional assessment of the student’s beliefs regarding simple and certain knowledge and justification by authority. As another

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 399

example, students who believe that knowledge in a domain is complex and dynamic are certainly not realists, but could be working from any of the three remaining positions. A skeptic would approach learning quite differently than a rationalist; therefore it is important for teachers to try to understand their students’ beliefs across the three dimensions. Therefore, if assessment tools for our model are to be used in class­rooms, they must help teachers combine the individual factor scores into indicators of positions. The factor scores in isolation from each other will tell teachers little about how their students’ beliefs regarding the nature of knowledge and knowing affect academic performance.

Finally, our conceptual model suggests that interventions can be tai­lored to specific positions. It is unlikely that a realist is going to move directly to a rationalist position, so pedagogies and interventions aimed at epistemic and ontological cognitive development should instead focus upon illustrating the complexity and interrelation of ideas within domains. Focusing upon helping students adopt more sophisticated beliefs regarding the nature of knowledge should be an early goal, because the model suggests this is the key to the next level of develop­ment. Interventions aimed at promoting evaluativist thinking may not be meeting realists’ needs, and instead should be saved for dogmatists and skeptics. Interventions designed to foster development beyond the dogmatist position may need to target both those students’ lack of faith in personal means of justification as well as their overreliance upon authority figures as a means of justification. Interventions for skeptics could be targeted in the opposite manner.

Educational implications of this study. Our findings also have implica­tions for educators. For example, in our study the elementary students had clear beliefs about knowledge and knowing in domains with which they were familiar, but had trouble thinking about domains they have not yet studied in school. It is interesting to speculate what happens when students are provided instruction in a domain they have not stud­ied before. Does initial exposure to an academic domain always result in naïve beliefs, regardless of the sophistication of the level of beliefs in similar domains? The evidence suggests that this is not the case. Politi­cal science, for example, is probably assimilated to previous learning about political history from social studies classes. In a study of high school students, most were able to comprehend a complex text about direct and representative democracy presenting a series of arguments between one group of political scientists favoring one, and one group favoring the other (Chambliss et al., 2007). Also remember that in our current study even elementary school students showed some under­standing that history knowledge had subjective elements, including

Using cognitive interviewing400

whether the American revolution was a “good” thing or not. This sug­gests that when exposing students to academic domains that are new to them, teachers may be able to help students by showing similarities between the new domain and familiar concepts and domains. In well­structured domains, for example, students may be more likely to adopt a more sophisticated position after being shown how solving problems using formulae in physics is similar to solving mathematical problems.

Further, we believe participants’ faith in the relative infallibility of math teachers is a double­edged sword. On the one hand, a focus upon the material, rather than on constantly evaluating the credibility of the teacher, may reduce extraneous cognitive load and allow students to more easily learn the material (van Merriënboer and Sweller, 2005). On the other hand, this belief may discourage students from seeking their own and perhaps alternative means of solving problems and understand­ing concepts. Students may initially benefit from their faith in teachers, but eventually they should be encouraged to find their own solutions and adopt a critical perspective upon knowledge and problem­solving in general. We do not mean to imply that students should doubt their teachers, but rather that they should be willing and able to critically examine what they learn inside and outside of the classroom.

Teachers may find our participants’ faith in computer websites dis­quieting. For our participants, the key issue seemed to be the perception that websites were created and vetted by multiple experts, causing stu­dents to treat them as academicians view peer­reviewed journal articles. Certainly the work in Internet literacy is informative here (see Eagleton and Dobler, 2006). However, we would suggest teachers stress the often solitary and unsupervised nature of building websites. Important dis­tinctions between websites and textbooks can be made by explaining how experts review textbooks for accuracy, and why a textbook writer’s credentials are often presented.

Finally, we find Hofer’s (2004b) ideas regarding the relation between beliefs about the nature of knowledge and learning strategies insight­ful. Our findings build upon Hofer’s work by highlighting that stu­dents’ beliefs about the nature of knowledge may contain more nuance than previously posited. In particular, two of our elementary school students appeared to understand that differences in perspective can lead to different evaluations of the relative positive or negative valence of historical events. This is not indicative of a truly “sophisticated” view regarding the nature of knowledge in history, but it may provide the means for scaffolding such beliefs. If teachers are able to expand upon this understanding of how perspective influences not only evalu­ations of valence but also the actual reporting of events, this may be a

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 401

powerful means of moving students’ view of history from a simple list of names, dates, and locations to a complex endeavor involving sourc­ing and triangulation of data (VanSledright, 2004). Likewise, teachers could target students’ ontology of mathematics and demonstrate that mathematical “facts” are often derived based upon certain assump­tions or axioms that can change depending upon context (i.e., Eucli­dean versus Boolean geometry). This more nuanced understanding can help illustrate how proofs that work under one set of assumptions do not work under others, one example of the complex and changing nature of knowledge in math.

Conclusion

In this chapter we have demonstrated how cognitive interviewing affords a different and informative perspective upon the measurement and conceptualization of students’ beliefs about the nature of knowledge and knowing. We have outlined how our model integrates educational, developmental, and philosophical perspectives regarding epistemic and ontological cognition. We have also provided evidence regarding the apparent sophistication of elementary and secondary school students, and how this could inform and even reform research into epistemic and ontological cognition. There are numerous educational interventions that can result from a better understanding of how students think about knowledge and knowing in various academic domains, but it is incum­bent upon researchers to base these interventions upon sound measure­ment and empirically supported conceptual models.

R EF ER ENCES

Alexander, P. A. (2006). What would Dewey say? Channeling Dewey on the issue of specificity of epistemic beliefs: A response to Muis, Bendixen, and Haerle (2006). Educational Psychology Review, 18(1), 55–65.

Alexander, P. A. and Dochy, F. J. R. C. (1995). Conceptions of knowledge and beliefs: A comparison across varying cultural and educational communi­ties. American Educational Research Journal, 32, 413–42.

Alexander, P. A., Murphy, P. K., Guan, J., and Murphy, P. A. (1998). How students and teachers in Singapore and the United States conceptualize knowledge and beliefs: Positioning learning with epistemological frame­works. Learning and Instruction, 8, 97–116.

Baxter Magolda, M. B. (2004). Evolution of a constructivist conceptualization of epistemological reflection. Educational Psychologist, 39(1), 31–42.

Beatty, P. C. and Willis, B. G. (2007). Research synthesis: The practice of cog­nitive interviewing. Public Opinion Quarterly, 71(2), 287–311.

Belenky, M. F., Clinchy, B. M., Goldberger, N. R., and Tarule, J. M. (1986). Women’s ways of knowing. New York: Basic Books.

Using cognitive interviewing402

Bendixen, L. D. and Rule, D. C. (2004). An integrative approach to personal epistemology: A guiding model. Educational Psychologist, 39, 69–80.

Boyer, E. L. (1987). College: The undergraduate experience in America. New York: Harper and Row.

Boyes, M. and Chandler, M. J. (1992). Cognitive development, epistemic doubt, and identity formation in adolescence. Journal of Youth and Adoles-cence, 21, 277–304.

Buehl, M. M. (2008). Assessing the multidimensionality of students’ epistemic beliefs across diverse cultures. In M. S. Khine (Ed.), Knowing, knowl-edge and beliefs: Epistemological studies across diverse cultures (65–112). New York: Springer.

Buehl, M. M., Alexander, P. A., and Murphy, P. K. (2002). Beliefs about schooled knowledge: Domain specific or domain general? Contemporary Educational Psychology, 27, 415–49.

Chambliss, M., Richardson, W., Torney­Purta, J., and Wilkenfeld, B. (2007). Improving textbooks as a way to foster civic understanding and engage­ment. College Park, MD: CIRCLE Working Paper, www.civicyouth.org (accessed April 25, 2008).

Chandler, M. J., Hallett, D., and Sokol, B. W. (2002). Competing claims about competing knowledge claims. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (145–68). Mahwah, NJ: Lawrence Erlbaum Associates.

Chinn, C. A. and Brewer, W. F. (1993). The role of anomalous data in knowl­edge acquisition: A theoretical framework and implications for science instruction. Cognition and Instruction, 19, 323–93.

Clarebout, G., Elen, J., Luyten, L., and Bamps, H. (2001). Assessing epistemo­logical beliefs: Schommer’s questionnaire revisited. Educational Research and Evaluation, 7, 53–77.

Conley, A. M., Pintrich, P. R., Vekiri, I., and Harrison, D. (2004). Changes in epistemological beliefs in elementary science students. Contemporary Edu-cational Psychology, 29, 186–204.

Creswell, J. W. (2003). Research design: Qualitative, quantitative, and mixed meth-ods approaches (2nd edn.). Thousand Oaks, CA: Sage Publishers.

DeVellis, R. F. (2003). Scale development: Theory and applications (2nd edn.), Vol. 26). Thousand Oaks, CA: Sage Publications.

Dole, J. A. and Sinatra, G. M. (1998). Reconceptualizing change in the cogni­tive construction of knowledge. Educational Psychologist, 33, 109–28.

Donald, J. (1990). University professors’ views of knowledge and validation process. Journal of Educational Psychology, 82, 242–9.

Eagleton, M. B. and Dobler, E. (2006). Reading the web: Strategies for internet inquiry. New York: Guilford Press.

Ericsson, K. A. (2006). Protocol analysis and expert thought: Concurrent ver­balizations of thinking during experts’ performance on representative tasks. In K. A. Ericsson et al. (Eds.), The Cambridge handbook of expertise and expert performance (223–42). Cambridge: Cambridge University Press.

Ericsson, K. A. and Simon, H. A. (1993). Protocol analysis: Verbal reports as data. Cambridge, MA: The MIT Press.

Flavell, J. H. (2004). Theory­of­mind development: Retrospect and prospect. Merill-Palmer Quarterly, 50, 274–90.

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 403

Frederiksen, N. (1984). Implication of cognitive theory for instruction in prob­lem solving. Review of Educational Research, 54, 363–407.

Gorsuch, R. L. (1983). Factor analysis (2nd edn.). Hillsdale, NJ: Lawrence Erl­baum Associates.

Glaser, B. (1992). Basics of grounded theory analysis. Mill Valley, CA: Sociology Press.

Glaser, B. and Strauss, A. (1967). The discovery of grounded theory. Chicago: Aldine.

Greene, J. A. (2007). A model of the development of epistemic and ontologic cogni-tion. Unpublished doctoral dissertation, College Park, MD: University of Maryland (August).

Greene, J. A. and Azevedo, R. (2007). A theoretical review of Winne and Had­win’s model of self­regulated learning: New perspectives and directions. Review of Educational Research, 77(3), 334–72.

(2008). The epistemic and ontologic cognitive development model: Formulation and testing. Paper presented at the American Educational Research Asso­ciation conference, New York, NY (March).

Greene, J. A., Azevedo, R., and Torney­Purta, J. (2008). Modeling epistemic and ontological cognition: Philosophical perspectives and methodological directions. Educational Psychologist, 43(3), 142–60.

Hallett, D., Chandler, M. J., and Krettenauer, T. (2002). Disentangling the course of epistemic development: Parsing knowledge by epistemic con­tent. New Ideas in Psychology, 20, 285–307.

Hmelo­Silver, C. E., Duncan, R. G., and Chinn, C. A. (2007). Scaffolding and achievement in problem­based and inquiry learning: A response to Kirschner, Sweller, and Clark (2006). Educational Psychologist, 42(4), 99–107.

Hofer, B. K. (2001). Personal epistemology research: Implications for learning and teaching. Journal of Educational Psychology Review, 13(4), 353–83.

(2002). Personal epistemology as a psychological and educational con­struct: An introduction. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (3–14). Mahwah, NJ: Lawrence Erlbaum Associates.

(2004a). Introduction: Paradigmatic approaches to personal epistemology. Educational Psychologist, 39, 1–4.

(2004b). Epistemological understanding as a metacognitive process: Thinking aloud during online searching. Educational Psychologist, 39, 43–56.

(2006). Beliefs about knowledge and knowing: Integrating domain specifi­city and domain generality: A response to Muis, Bendixen, and Haerle (2006). Educational Psychology Review, 18(1), 67–76.

(2008). Personal epistemology and culture. In M. S. Khine (Ed.), Knowing, knowledge and beliefs: Epistemological studies across diverse cultures (3–22). New York: Springer.

Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learn­ing Review of Educational Research, 67, 88–140.

Using cognitive interviewing404

Karabenick, S. A., Woolley, M. E., Friedel, J. M., Ammon, B. V., Blazevski, J., Rhee Bonney, C., et al. (2007). Cognitive processing of self­report items in educational research: Do they think what we mean? Educational Psycholo-gist, 42(3), 139–51.

King, P. M. and Kitchener, K. S. (1994). Developing reflective judgment: Under-standing and promoting intellectual growth and critical thinking in adolescents and adults. San Francisco: Jossey­Bass.

(2004). Reflective judgment: Theory and research on the development of epistemic assumptions through adulthood. Educational Psychologist, 39, 5–18.

Kitchener, R. (2002). Folk epistemology: An introduction. New Ideas in Psy-chology, 20, 89–105.

Kuhn, D. (2005). Education for thinking. Cambridge, MA: Harvard University Press.

Kuhn, D., Cheney, R., and Weinstock, M. (2000). The development of episte­mological understanding. Cognitive Development, 15, 309–28.

Kuhn, D. and Weinstock, M. (2002). What is epistemological thinking and why does it matter? In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing (121–44). Mahwah, NJ: Lawrence Erlbaum Associates.

Muis, K. R., Bendixen, L. D., and Haerle, F. C. (2006). Domain­generality and domain­specificity in personal epistemology research: Philosophical and empirical reflections in the development of a theoretical framework. Educational Psychology Review, 18, 3–54.

Murphy, P. K. (2005). Toward a model of topic knowledge and belief change. Paper presented at the annual conference of the Southwest Consortium for Inno­vative Psychology, Las Vegas, Nevada (November).

(2007). The eye of the beholder: The interplay of social and cognitive com­ponents in change. Educational Psychologist, 42(1), 41–53.

Murphy, P. K. and Alexander, P. A. (2004). Epistemological threads in the fabric of conceptual change. Invited paper presented at the 4th European Sympo­sium on Conceptual Change, Delphi, Greece (May).

Murphy, P. K., Alexander, P. A., Greene, J. A., and Edwards, M. N. (2007). Epistemological threads in the fabric of conceptual change. In S. Vosniadou (Ed.), Reframing the conceptual change approach in learning and instruction (105–22). New York: Elsevier Press.

Neber, H. and Schommer­Aikins, M. (2002). Self­regulated science learning with highly gifted students: The role of cognitive, motivational, epis­temological, and environmental variables. High Ability Studies, 13(1), 59–74.

Patton, M. Q. (1990). Qualitative evaluation and research methods (2nd edn.). Newbury Park, CA: Sage Publications.

Perry, W. G. (1970). Forms of ethical and intellectual development in the college years: A scheme. New York: Holt, Rinehart and Winston.

(1999). Forms of ethical and intellectual development in the college years: A scheme. San Francisco: Jossey­Bass.

Plantinga, A. (1993). Warrant: The current debate. New York: Oxford University Press.

J. A. Greene, J. Torney­Purta, R. Azevedo, and J. Robertson 405

Pollock, J. L. and Cruz, J. (1999). Contemporary theories of knowledge (2nd edn.). New York: Rowman and Littlefield.

Qian, G. and Alverman, D. (1995). Role of epistemological beliefs and learned helplessness in secondary school students’ learning science concepts from text. Journal of Educational Psychology, 87(2), 282–92.

Richardson, W. K. (2006). Combining cognitive interviews and social science surveys: Strengthening interpretation and design. In K. C. Barton (Ed.), Research methods in social studies education: Contemporary issues and perspec-tives (159–81). Greenwich, CT: Information Age Publishing.

Schommer, M. (1990). Effects of beliefs about the nature of knowledge on comprehension. Journal of Educational Psychology, 82, 398–504.

(1993). Epistemological development and academic performance among secondary students. Journal of Educational Psychology, 85(3), 406–11.

Schommer, M., Calvert, C., Gariglietti, G., and Bajaj, A. (1997). The develop­ment of epistemological beliefs among secondary students: A longitudinal study. Journal of Educational Psychology, 89, 37–40.

Schommer, M., Crouse, A., and Rhodes, N. (1992). Epistemological beliefs and mathematical text comprehension: Believing it is simple does not make it so. Journal of Educational Psychology, 84, 435–43.

Schommer­Aikins, M. (2004). Explaining the epistemological belief sys­tem: Introducing the embedded systemic model and coordinated research approach. Educational Psychologist, 39, 19–29.

Schommer­Aikins, M., Brookhart, S., Hutter, R., and Mau, W. (2000). Under­standing middle­school students’ beliefs about knowledge and learning using a multidimensional paradigm. The Journal of Educational Research, 94, 120–7.

Schommer­Aikins, M. and Easter, M. (2006). Ways of knowing and episte­mological beliefs: Combined effect on academic performance. Educational Psychology, 26, 411–23.

Schraw, G. J. and Olafson, L. J. (2008). Assessing teachers’ epistemologi­cal and ontological worldviews. In M. S. Khine (Ed.), Knowing, knowl-edge and beliefs: Epistemological studies across diverse cultures (25–44). New York: Springer.

Schraw, G., Bendixen, L. D., and Dunkle, M. E. (2002). Development and validation of the epistemic belief inventory (EBI). In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowl-edge and knowing (261–76). Mahwah, N J: Lawrence Erlbaum Associates.

Slotta, J. D. and Chi, M. T. H. (2006). Helping students understand challeng­ing topics in science through ontology training. Cognition and Instruction, 24(2), 261–89.

Stodolsky, S. S., Salk, S., and Glaessner, B. (1991). Student views about learn­ing math and social studies. American Educational Research Journal, 28, 89–116.

Strauss, A. and Corbin, J. (1998). Basics of qualitative research: Techniques and procedures for developing grounded theory (2nd edn.). Thousand Oaks: Sage.

Sullivan, W. M. and Rosin, M. S. (2008). A new agenda for higher educa-tion: Shaping a life of the mind for practice. San Francisco: Jossey­Bass.

Using cognitive interviewing406

Uebersax, J. S. (2006). Likert scales: Dispelling the confusion. Statistical meth­ods for rater agreement website. Available at: http://ourworld.compuserve.com/homepages/jsuebersax/likert2.htm (accessed April 19, 2008).

VanSledright, B. A. (2004). What does it mean to think historically … and how do you teach it? Social Education, 68(3), 230–3.

VanSledright, B. A. and Limón, M. (2006). Learning and teaching social studies: A review of cognitive research in history and geography. In P. A. Alexander and P. H. Winne (Eds.), Handbook of Educational Psychology (2nd edn.), (545–70). Mahwah, NJ: Lawrence Erlbaum Associates.

van Merriënboer, J. J. G. and Sweller, J. (2005). Cognitive load theory and complex learning: Recent developments and future directions. Educational Psychology Review, 17(2), 147–77.

Williams, M. (2001). Problems of knowledge: A critical introduction to epistemol-ogy. New York: Oxford University Press.

Willis, G. B. (2005). Cognitive interviewing: A tool for improving questionnaire design. Thousand Oaks, CA: Sage Publications.

Wood, P. and Kardash, C. (2002). Critical elements in the design and analy­sis of studies of epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (231–60). Mahwah, NJ: Lawrence Erlbaum Associates.

Part IV

Teachers’ personal epistemology and its impact on classroom teaching

409

13 Epistemological resources and framing: a cognitive framework for helping teachers interpret and respond to their students’ epistemologies

Andrew Elby and David HammerUniversity of Maryland at College Park

Previously, we have argued that an account of personal epistemologies based on epistemological resources shows generativity and explanatory power, especially for understanding variability in a student’s behavior. In this chapter, we argue that a resources framework is generative for instruction and is therefore worth teaching to teachers. Using for illus­tration a case study of middle school Earth science students learning about the rock cycle, we argue that the resources framework: (1) pre­dicts the existence of coherent networks of resources that correspond to what teachers can recognize, and what novice teachers can learn to rec­ognize, in students’ approaches to learning; (2) invites close attention to context when evaluating whether a given student utterance or behavior reflects a productive stance toward knowledge, leading to more nuanced assessments of the student’s approach to learning; and (3) provides guidance about how to foster epistemological sophistication over both short and long time scales. To support these points, we first extend the resources framework to address a challenge it presents: epistemological resources are rarely apparent in isolation. Instead, the main observable grain­size of student epistemologies corresponds to an epistemological frame, a locally coherent activation of a network of resources that may look like a stable belief or theory. A particular epistemological resource, we argue, can play different roles in different frames, and this feature of our framework has instructional implications.

Introduction

A growing body of research on personal epistemologies contends that how students understand the nature of knowledge, knowing, and learning affects how they reason and learn (Hammer, 1994; Hofer

Epistemological resources and framing410

and Pintrich, 1997; Hogan, 1999; May and Etkina, 2002; Sandoval, 2005; Schommer, 1993; Schommer et al., 1992).1 Readers of this vol­ume will not find it controversial to conclude that preschool to grade twelve teacher education and professional development should address the notion of personal epistemology.

In this chapter, we argue that the resources perspective is particularly useful and generative for preschool to grade twelve teachers. To make this argument, we first review and extend the resources perspective to include epistemological framing (Hammer et al., 2005; Redish, 2004), which corresponds closely to the phenomena of coherent personal epis­temologies documented in the literature (Hofer and Pintrich, 1997). We then offer three reasons why this perspective is particularly produc­tive for teachers.

Epistemological resources and framing

We first review our cognitive framework for describing personal epis­temologies in terms of epistemological resources. Then we extend that framework to account for the coherent, belief­like epistemologies that students often display (diSessa et al., 2002; Hammer, 1994; King and Kitchener, 2002; Lising and Elby, 2005). Our framework, however, predicts that epistemological coherence is often local rather than glo­bal, i.e., that students manifest different coherent epistemologies in dif­ferent contexts.

Review of epistemological resources

Cognitive resources are fine­grained knowledge elements that a student possesses, the activation of which depends on context (diSessa, 1993; Hammer, 2000). To see how resources differ from the “beliefs” or “conceptions” that researchers often attribute to students (Carey, 1992; Hewson, 1981; McCloskey, 1983), consider an example: many students expect that an ice cube wrapped in cloth melts more quickly than one left out in the air. One interpretation of the students’ thinking would be to attribute a misconception that some materials, such as blankets, jackets, and gloves, are inherently warm. But ask those same students to think about removing a hot pan from the oven, and they would view that same cloth as something that can protect their hands by “blocking” the

1 Like Schommer­Aikins (2004), and unlike Hofer and Pintrich (1997) and others, we include views about learning in our definition of personal epistemology. We have both theoretical (Elby, 2009) and practical reasons for this choice, as discussed below.

Andrew Elby and David Hammer 411

heat. The resources framework invites an explanation of these different patterns of reasoning in terms of the context­dependent activation of cognitive resources. From their experiences associating blankets and gloves with warmth, students may have formed mini­generalizations about cause and effect – that softness means warmth, perhaps. The cloth­wrapped ice cube triggers this resource. Other contexts, however, turn on different resources. When reasoning about the cloth and hot pan, students may activate their intuitive idea of blocking (diSessa, 1993) but not softness means warmth.

In our framework, context­sensitive activation of fine­grained knowl­edge elements also describes students’ cognition concerning the nature of knowledge, knowing and learning (Louca et al., 2004). A college student we described in a previous paper illustrates what we mean (Hammer et al., 2005). When asked how he prepared for his physics test, Louis said that he “studied every word of those homework solu­tions … I was memorizing the book, too.” This response reflects a view of knowledge as something absorbed from an authoritative source. By contrast, talking about his strategies for tutoring other students, Louis said, “what I like to do is build on what they already know instead of introducing a totally new concept,” reflecting a view of knowledge as something constructed out of prior knowledge.

Louis’s variability can be understood as arising from the different contexts activating different resources, in this case resources for under­standing the nature of knowledge and how an individual comes to have it. Note that “different contexts” includes but is not confined to dif­ferent disciplinary domains (Muis et al., 2006). As a physics student, Louis activated knowledge as propagated stuff, a resource for understand­ing knowledge as passed from a source to a recipient. This attribu­tion does not contradict an explanation of Louis’s behavior in terms of long­established habits and expectations about school. Those expecta­tions and habits, in our framework, are likely to activate and to be acti­vated by a network of epistemological resources that includes knowledge as propagated stuff. As a tutor, by contrast, Louis activated knowledge as constructed, a resource for understanding knowledge as built from other knowledge. That activation may have occurred because Louis was focused on his tutee’s knowledge rather than his own, or because he was familiar with the academic material.

We need to clarify how the knowledge as propagated stuff and know ledge as constructed resources differ from the transmissionist and constructivist beliefs often discussed in the personal epistemology literature (Hofer and Pintrich, 1997). In earlier work (Hammer and Elby, 2002), we posited knowledge as propagated stuff as a little piece of knowledge about

Epistemological resources and framing412

knowledge that enables a kindergartener, when asked how she knows what is for dinner, to answer, “Because Daddy told me!” This epistemo­logical resource enables her in some circumstances to view knowledge as a kind of thing that can be passed from source to recipient, even if she cannot articulate this view or apply it consistently. Similarly, the kinder­gartener possesses knowledge as constructed, a resource that enables her to view knowledge as something that can be figured out. For instance, when her birthday arrives and her mother walks in holding something behind her back, the kindergartener might squeal, “A present!” When asked how she knows, she might say, “It’s my birthday and I saw you holding something – I figured it out.” We use a kindergartener to emphasize that these resources are not beliefs; a kindergartener’s “transmissionist” and “constructivist” views almost certainly lack the articulateness, robust­ness, or stability of beliefs. However, these and other epistemological resources can form coherent, stable networks that constitute the episte­mological beliefs observed in older children and adults.

Before discussing the role these networks play in our framework, we briefly mention two other characteristics of epistemological resources. First, in both children and adults, the activation of a resource can occur beneath the subject’s awareness. When we say Louis activated knowl-edge as propagated stuff, we do not imply that he consciously chose to use that resource. Second, these resources cut across the categories of “knowledge” and “learning.” For instance, knowledge as propagated stuff, functioning as an element in a cognitive model of personal epistemolo­gies, contributes to a view of knowledge as a kind of shareable stuff and to a view of learning as receiving information. We include views about learning within our definition of personal epistemology not for disci­plinary or other a priori reasons but because, in the cognitive models we are building, views about learning are inextricably entangled with views about knowledge and knowing (Elby, in press).

Epistemological framing

We do not view personal epistemologies as haphazard and incoherent. Rather, we expect local coherences, including belief­like consistencies and stabilities, within a given context (Hammer, 1994) and sometimes across contexts (Lising and Elby, 2005). We now explain how the resources framework accounts for epistemological coherence.

Framing. Epistemological resources do not generally turn on or off in isolation; the activation of one resource can promote or inhibit the

Andrew Elby and David Hammer 413

activation of others. We lack a precise developmental story, but we sub­scribe to a dynamic systems view (Thelen and Smith, 1994) of develop­ment as a progressive construction of patterns of resource activation.

In some cases, a network of activated resources will be unstable and perhaps turn off quickly. In other cases, the network of resources will be locally coherent, by which we mean the activations of the individual resources are mutually reinforced by each other and by contextual cues and other features of the environment, leading to stability. A locally coherent activation of a network of epistemological (and perhaps other) resources is an epistemological frame (Hammer et al., 2005).

Louis, for example, recounted succeeding in science classes by memorizing:

Because in all, especially in my like chemistry classes, the way I did well on the exam is like flash cards of different reactions and memorize it and the better I memorized it the better I did on those exams. (Hammer et al., 2005)

He assumed that physics would be similar. So, for him, features of the context as filtered through his previous experiences tended to activate knowledge as propagated stuff, accumulation (a resource contrib­uting to a view of learning as accumulating information), and other resources. Resources in this network probably cue each other, too. For instance, thinking of knowledge as propagated stuff, passing from a source to a recipient, coheres with thinking of learning as accumulation of that stuff. These mutual cueings produce local coherence: Louis in phys­ics class is stable in framing learning as memorization. (Notice that this “memorization” frame includes views about knowledge and views about learning, deeply entangled.)

Redish (2004) proposed the term “epistemological frame” to con­nect the study of personal epistemologies to the notion of framing from anthropology and sociolinguistics. In Tannen’s (1993) sociolinguistic analyses, framing is the set of expectations an individual brings to a social situation, expectations that affect what she notices and how she thinks to act. Roughly speaking, in Tannen (1993), a framing is a per­son’s generally tacit answer to the question “What sort of activity is this?” For instance, a father at his child’s soccer game can frame his activity as rooting for a sports team or as nurturing children. Those two different framings lead the father to notice different things (e.g., who is winning versus who is happy) and to behave in different ways (e.g., partisan cheering and jeering versus general encouragement). Tannen’s (1993) conceptualization of framing corresponds in our perspective to an individual’s forming of a locally coherent activation of resources (Hammer et al., 2005). In particular, epistemological framing is the

Epistemological resources and framing414

cognitive activity underlying a student’s sense of “what is going on here” with respect to knowledge.

The soccer example illustrates a crucial point about framing, namely context dependence and shifting. The father might frame his child’s soccer game differently in different contexts, such as an unscored scrimmage versus the league championship, and he might shift frames in response to contextual cues, such as his child getting hurt or other parents reminding him it is just a game. Similarly, although Louis ini­tially framed tutoring much differently from the way he framed learn­ing physics, an intervention helped him enact a shift (Hammer et al., 2005). When Louis visited office hours after failing the first midterm exam, the professor advised him “When you study, try to explain it to a ten­year­old.” In Louis’ recollection, this advice led him to approach the rest of the course in a different, more sense­making way, more like the way he approached tutoring. His performance improved dramati­cally, and he ended up with a B+.

We have evidence that supports Louis’s explanation of his improve­ment (Hammer et al., 2005). First, in an interview with a researcher who was never his professor, Louis not only stated how he changed his study strategies but also gave specific examples, including an anal­ogy he generated for understanding electric charges moving through wires based on trucks driving on roads. Second, Louis contacted the professor after the semester and spontaneously attributed his success specifically to the professor’s intervention as opposed to the professor’s lectures, labs, and so on. That success, Louis recounted, extended to other courses; his grade point average (GPA) had increased, which he again attributed to a general shift in how he approached learning. For these reasons, we find it plausible that the intervention helped Louis epistemologically reframe his approach to learning, activating a locally coherent network of resources – a frame – similar to the one he acti­vates while tutoring, a frame that includes knowledge as constructed and related resources.

Regrettably, such one­time interventions rarely work so well. One possible explanation for why it helped Louis is that he had already formed that productive frame in another context.

Framing is believing. From a resources perspective, the coherent epis­temological “beliefs” that researchers observe in their subjects reflect stable epistemological frames. That stability could arise in a variety of ways. Elsewhere we described three mechanisms of stability: contex­tual, deliberate, and structural (Hammer et al., 2005).

Andrew Elby and David Hammer 415

Contextual: the stability relies on persistent cueing by features of the con­text. For example, a student might adopt an active approach to learning, treating knowledge as personally constructed, in a setting where reformed instruction and curricular materials prompt and scaffold that framing. But the student might shift back to rote learning, treating knowledge as information received from authority, when those cues are removed.

Deliberate: the stability of the epistemological frame relies on the indi­vidual’s deliberate attention toward maintaining a consistent stance. Louis, for example, said his choice to treat physics knowledge as built out of everyday ideas competed against his established routines. At first, he had to keep reminding himself to explain ideas in simple terms rather than revert to rote memorization, an example of epistemological voli­tion (Bendixen and Rule, 2004; Corno, 1993) relying on meta­cognitive monitoring in an epistemological context (Hofer, 2004).

Structural: a coherent pattern of resource activations can become stable over time, as it is used and reused. In that case, the network of resources turns on as a compiled, robust unit of cognitive structure; it becomes a resource in its own right. Louis’ success in physics class may have stemmed in part from his needing to keep his “constructiv­ist” framing stable by deliberate means for a comparatively short time before it compiled in that context into a structurally stable cognitive unit. A contextually and/or deliberately stable epistemological frame can develop, through repeated use, into a structurally stable frame.

In summary, from the resources perspective, an epistemological “belief” professed or enacted in a given context – a physics class, an interview about controversial topics, an epistemological survey given in psychology class – corresponds to an epistemological frame that may be more stable or less stable by any of these three mechanisms. Only the last mechanism, structural stability, corresponds to the use of “belief” in the literature. For example, epistemology researchers typically describe a belief as something an individual “holds” (Hofer and Pin­trich, 1997). When an epistemological frame is only contextually stable, however, the observed pattern of epistemological thought persists only in particular contexts, making it problematic to speak of that pattern as belonging entirely to the individual.

Methodological implications. By this account of epistemologies, research­ers cannot assume that the “belief” a subject exhibits in one context is a global or even domain­specific epistemological belief. Only by prob­ing the subject in different contexts doing different activities, within or across domains, can we determine the generality of the “belief.”

Epistemological resources and framing416

Tweaking the experimental conditions in various ways would also provide insights about the mechanism underlying whatever degree of stability is observed. For instance, changing the location of an inter­view from the physics building (with the interviewer introduced as a physicist) to the education building (with the interviewer introduced as a psychologist), while otherwise keeping the protocol the same, could uncover contextual stabilities and instabilities. Changing the ordering of the questions, from a random jumble to a more ordered progression that invites coherence, could shed light on deliberate stabilities and instabilities.

Our framework also suggests the need for new models of epistemolog­ical change. In Bendixen and Rule’s (2004) model, which synthesizes recent theoretical and empirical work, such change generally involves epistemic doubt and/or volition and results in new epistemological beliefs. In our model of Louis’ epistemological change, by contrast, vol ition plays a role, but the end result was not a new epistemological belief; it was the redeployment of an old “belief.” More generally, in our framework, much epistemological development and change consists of the co­activation and stabilization of epistemological resources the stu­dent already possesses.

Connections to teachers’ professional knowledge

We have argued that a resources­based account draws on and may con­tribute to teachers’ professional knowledge (Hammer and Elby, 2003). Our experiences as instructors informed our development of a resources­based perspective. In particular, those experiences concerned students’ approaches to learning and how they may shift, depending on how they understand (frame) the task.

Here, we refine that argument to focus on framing, following Redish (2004). To review, an epistemological frame, at the level of experi­ence, is an individual’s sense of “what is going on here” with respect to knowledge. At the level of cognitive structure, it is a locally coherent activation of a network of epistemological and associated resources. We argue that

(1) epistemological frames correspond to what teachers can recognize, and novice teachers can learn to recognize, in students’ approaches to learning;

(2) the possibility that a given resource can participate in multiple frames invites close attention to context when evaluating whether a given student utterance or behavior reflects a productive stance

Andrew Elby and David Hammer 417

toward knowledge, leading to more nuanced assessment of the stu­dent’s approach to learning; and

(3) the resources perspective provides guidance about fostering episte­mological change over both short and long time scales.

Rather than draw again on our own teaching experiences, we develop our arguments with respect to a case study from an eighth­grade sci­ence class taught by Jessica Phelan (Phelan, 2006; also Rosenberg et al., 2006). To be clear, we are not empirically “proving” our claims; we are illustrating them with an example from a case study.

Epistemological frames correspond to what teachers can recognize in students’ approaches to learning

We illustrate this first claim with an example of Ms. Phelan’s diagnosis of student difficulties as they worked to develop a model of the rock cycle.

The students had completed worksheets addressing the formation and properties of igneous, sedimentary, and metamorphic rocks, but had received no formal instruction on the rock cycle, the physical proc­esses by which those three kinds of rocks turn into each other. Ms. Phe­lan split the twenty­five­person class in half and instructed each group to make a model of the rock cycle.

One group sat in a circle on the floor in the hallway. This is how they began their work.

lisa: I need to get out the papers.bethany: OK, so what is the rock cycle?ben: Well, it starts out as an igneous rock. Right? And then it um – and then

it like –johanna: An igneous rock forms, weathering occurs.bethany: OK, wait … Igneous rock.johanna: [sing­song voice, reciting] Igneous rock forms, weathering occurs

… weathering.ryan: I have all of my sheets if we’re allowed to go back in the room.bethany: OK, what happens next? Lisa? [Lisa has her papers out.]lisa: Igneous rock … it forms from magma (inaudible) lava. First we have to

start with the plates running into each other, and the lava going up.tracy: It’s either erosion (inaudible) or sediments are formed.

While Lisa is speaking several other students chime in with comments about the “deposits.” Tracy decides erosion comes first, and tells Beth­any to record that.

ben: The deposit goes through erosion, and settles at the bottom of the sea.lisa: First we have to start with uh first we have to start with the plates –

underground.

Epistemological resources and framing418

bethany: Oh wait, so what happens?tracy: (inaudible) plates underground.bethany: So, the Teutonic (sic) plates move, and –lisa: Yeah, that’s the very beginning.bethany: OK. [Starting to write] Teutonic plates …

The conversation stops as Bethany writes, and she motions for the group to continue: “You guys can talk while we’re writing.” At this moment, Ms. Phelan comes out to check on their progress, in time to hear Bethany reading back a summary to the group of what they had accomplished thus far:

ben: And then, after that happens, the sediment goes to the bottom of the ocean, and then it compresses to form …

bethany: [reading] So, the Teutonic plates move and create rock, and then I have the igneous rock forms. Is that wrong? [Pause. Ben and Ryan both start to say something but hesitate.]

lisa: No, there’s something. There was like de­desa­whatever.johanna: Deposit (inaudible).lisa: Yeah, deposition.tracy: We’re not there yet.ben: The deposit comes after that.

Rosenberg et al. (2006), analyzing the conversation line­by­line, argued that the students were organizing their efforts around information retrieval rather than sense­making, focusing on terminology, and creat­ing a formal ordering rather than a causal story. Their analysis included the conversation that took place before Ms. Phelan entered the hallway, but there is evidence in the segment she overheard: the students discuss terminology (Teutonic plates, deposition) and temporal ordering (“So, the Teutonic plates move and create rock, and then I have the igneous rock forms. Is that wrong?” and “The deposit comes after that”), but they do not consider causal mechanism. Instead of coding or otherwise reducing the data, Rosenberg et al. (2006) argue from the substance of the students’ utterances, considering and rebutting plausible alternate interpretations, and presenting the entire transcript that they analyzed so that readers can judge for themselves the validity of the interpreta­tions. They then modeled the students’ framing as a locally coherent activation of five epistemological resources including knowledge as prop-agated stuff and accumulation. (The other three resources in their model are information, corresponding to the view that knowledge comes in the form of specific terminology and phrasings; ordered list, the cognitive machinery for understanding the knowledge form they are producing; and ordering, a resource for understanding the epistemic activity of cre­ating an ordered list.)

Andrew Elby and David Hammer 419

Overhearing part of this conversation, Ms. Phelan arrived at an inter­pretation similar to that of Rosenberg et al. (2006). Here is her entire intervention:

ms. phelan: Can I make a suggestion?bethany: Yeah.ms. phelan: You’re looking at a lot of papers and using a lot of words that you

don’t know what they mean.gustavo: Sure we do (?). [Ryan laughs.]ms. phelan: And if you’re doing that, for your model, it’s not going to be very

good. So, I want to start with what you know, not with what the paper says.

Note that she says nothing about rock concepts or the group dynam­ics; her intervention consists entirely of pointing out to students how they are framing the activity and suggesting that they frame it differ­ently. Specifically, she diagnoses the students to be framing knowledge as words to be retrieved from worksheets without comprehension, and she responds by suggesting that they reframe their activity to “start with what you know.” She implicitly expects that the students can keep their (re)framing stable by deliberate means, by monitoring their think­ing to make sure they are starting with what they know and not just regurgitating vocabulary. In brief, Ms. Phelan aimed her intervention at students’ approach to knowledge generation, and she expected that students could shift to a different approach, from one we would model as centered around knowledge as propagated stuff to one we would model as centered around knowledge as constructed. Her reflections on the les­son confirm this interpretation:

Hearing this bit of conversation made me think the students were not making sense out of the problem … They were using lots of big words, without under­standing their meanings. This wasn’t what I had in mind with the assignment! And it wasn’t close to what these students were capable of doing. I had seen them use reasoning skills to solve problems before … I felt that they had every­thing they needed to solve the problem of the rock cycle in their heads. It was just a matter of getting them to really think. (Phelan, 2006)

We continue the case study below. But first, we discuss two ways in which Ms. Phelan’s diagnosis and intervention align with the notion of epistemological frames: her expectation that students have multiple epistemological stances available to them, and the grain size of what she attended and responded to in her students’ thinking.

Expectation of epistemological variability. Ms. Phelan’s diagnosis and inter­vention reflect an intuitive understanding of student epistemologies

Epistemological resources and framing420

that, like the resources perspective, expects students to be capable of treating knowledge and learning in multiple ways, such as focusing on worksheets or “starting from what you know.” From attending to her students’ thinking in this episode and in previous classes, she knows her students’ abilities and some of the factors that may affect their reasoning.

We note, too, that this interpretation is continuous with laypersons’ everyday expertise about other people’s mind states. People with no teaching experience can recognize when a friend is getting bogged down in terminology and technical details rather than stepping back and making sense of the situation. Consider, for example, this discussion between two introductory college physics students answering a ques­tion about light and shadows. The four­person group which included Jan and Veronica was videotaped with no researcher present (Lising and Elby, 2005):

j a n : So the light is like that and these are the rays, and the vector points that way … All the rays are going like this. So, it’s kind of like polarized.

v e ron ic a : Mmm, not really. It’s just, well, it’s just … you’re making it, you’re trying to make it more difficult. It’s just, the light goes out … [common­sense explanation follows].

Veronica’s interpretation of her fellow student’s reasoning is similar to Ms. Phelan’s and Rosenberg et al.’s (2006) interpretation of the eighth graders: she sees Jan as taking a counterproductive stance toward the ideas in the lab, and she expects it is possible for Jan to adopt a different stance.

Grain size of the epistemological “unit.” Elsewhere we have argued that a resources­based account of student epistemologies aligns with teach­ers’ professional knowledge (Hammer and Elby, 2003), giving exam­ples from our own teaching. The refinement of that argument offered here applies to those examples as well: what a teacher can see of student epistemologies in the classroom corresponds to the grain size of an epis­temological frame rather than of an isolated epistemological resource. In Ms. Phelan’s (and our) interpretation, the students are framing their epistemic activity as retrieving information from authority. Although as researchers we view that framing as comprised of resources at a finer grain size, as teachers we notice the pattern (framing) as a whole.

The grain size of what Ms. Phelan notices and responds to aligns equally well with the notion of an epistemological belief (Hofer and Pintrich, 1997). Ms. Phelan’s interpretation and resulting interven­tion, however, align more closely with frames than beliefs. Rather than

Andrew Elby and David Hammer 421

thinking of students as holding beliefs or theories about the nature of knowledge, we think of students as framing epistemic activities in differ­ent ways at different times. There is coherence to their framing, but it may be local to the context. This was Ms. Phelan’s interpretation, and she acted accordingly: rather than eliciting and confronting an epis­temological belief that the relevant knowledge comes from authority (the worksheets), she simply noted the difference between focusing on “what the paper says” and “start[ing] from what you know.” The stu­dents’ adoption of a particular epistemological stance may be local to the circumstances, and a teacher making that diagnosis might expect that interventions aimed at changing those circumstances could result in substantial shifts. As we recount below, that is what happened with Ms. Phelan’s students.

Epistemological resources and views about learning. Finally, we note our practical reason for including views about learning along with views about knowledge and knowing in our definition of personal epistemolo­gies: what we see students doing in class, at the grain size of framing, almost always involves aspects of both. Because we are ultimately inter­ested in how students approach knowledge and learning in situations such as these, it serves us to treat knowledge and learning together as part of epistemic cognition. If our aims were otherwise – for example, if we were ultimately interested in students making progress as philoso­phers – we might find it more useful to tease apart students’ views about knowledge and knowing from their views about learning.

These arguments have implications for teacher education and profes­sional development. We urge teacher educators to focus not just on the default epistemological “beliefs” students most commonly exhibit, but also on the multiplicity of approaches to knowledge and learning that a student can display when nudged appropriately. And we urge that these discussions about epistemology be grounded in real classroom data, such as Ms. Phelan’s case study.

We now consider how teachers assess the productivity of student epistemologies.

The framing perspective invites attention to context in assessing whether students are taking a productive stance toward knowledge

We argue here that, because a given resource connected to a particular behavior can participate in multiple frames, the resources perspective

Epistemological resources and framing422

invites attention to how students frame the overall activity in which a particular behavior is embedded, leading to more nuanced judgments about the epistemological productivity of that behavior.

In previous work (Elby and Hammer, 2001) we argued that a resources­based view belies simple characterizations of epistemological sophistication. For example, while viewing scientific knowledge as ten­tative is often sophisticated, in some contexts it is counterproductive, such as in popular treatments of evolution as “just a theory” or in off­hand dismissals of global warming as “controversial.” In this and other ways, our notion of epistemological sophistication aligns with the eval­uativist epistemological level discussed by Kuhn (Kuhn, 1991; Kuhn and Weinstock, 2002).2

The notion of epistemological frames as locally coherent activations of finer­grained resources enables us to refine and extend our arguments about the nuanced, contextual nature of epistemological sophistication, and this refined argument has pedagogical implications. The key idea is that different epistemological frames draw upon overlapping rather than disjoint sets of epistemological resources. So, a particular episte­mological resource such as knowledge as propagated stuff, which can drive a specific behavior such as retrieving information from worksheets, par­ticipates in multiple epistemological frames. Consequently, when judg­ing the sophistication of a particular epistemological resource and of the behaviors it helps to drive, we must attend to the overall frame of which it is a part. A given behavior such as information retrieval may be epis­temologically productive or unproductive, depending on how students frame the activity in which it is embedded.

Illustration from the rock cycle discussion. To illustrate the importance of this idea for instruction, we return to the rock cycle discussion. After the teacher’s suggestion to “start with what you know, not with what the paper says,” the students shift how they frame the activity. Still, they continue to retrieve information from the paper (worksheets), which one might construe as not following the teacher’s advice. However, they now use the retrieved information in a different, more epistemologically productive way.

We pick up the conversation several minutes after Ms. Phelan’s intervention. Gustavo and Ben have been talking about how layers of

2 But we contest Kuhn’s (1991) stage­based conceptualization. Kuhn and Weinstock’s (2002) epistemological “levels” allow for more intrapersonal developmental variation and context dependence, though perhaps less than the framing perspective would predict.

Andrew Elby and David Hammer 423

sedimentary rock “[go] up and up” until there is a “big sedimentary rock” that is compressed by all the layers. Soon, the group turns to the question of how the sedimentary rock layers become metamorphic:

l i s a : We need heat and pressure … Where [do] we get more heat and pres­sure from?

j oh a n n a : [singsong, reciting] A metamorphic, a metamorphic rock forms from heat and pressure … applied to [pause] a rock. [laughter] To any rock. To any rock, and it makes it – it changes the grains, because the heat and pressure changes the crystal structure, the texture, the, appear­ance, an – and so – and so because – because it’s an immen­immense amount of heat or pressure, it can change any rock, and like the grains are changed, so it’s foliated or non­foliated.

While Johanna is speaking, Lisa and Bethany try to interrupt to ask where the heat and pressure come from. Ryan tries to answer, “From the lava of course,” and Lisa refines her question.

l i s a : Pressure I can understand where from, but the heat?b e t h a n y : Wait, where does it get heat and pressure?r ya n : From the earth’s core.j oh a n n a : From well – ‘cause it – ‘cause it’s underground, so it’s closer to the

core.l i s a : How did it get underground?

Several students speak at once to answer, with Bethany saying, “there’s layers and layers and layers on top” and Ryan saying that they “make it go closer to the core, the pressure from all the other layers.” Johanna asks, “Couldn’t it have started underground?”

After another brief exchange, the students wait for Tracy (the new recorder) to finish writing, and she says that the last thing she has is “Sediment rock is formed.” Bethany asks, “What happens after that?” and begins to read.

Bethany [reading from a worksheet]: Immense heat and pressure, deep beneath the earth’s surface, change the rock into a metamorphic rock.

Ben and Gustavo immediately return to explaining how the heat arises. They say it is because “so many layers makes it closer to the magma” at the core. Johanna points out the inconsistency: the group’s account of sedimentary rock layers building up from the bottom does not explain how the resultant sedimentary rock gets closer to the Earth’s core.

In some sense the students did not follow all of Ms. Phelan’s advice; both Johanna and Bethany read from the worksheets, and the group continues to use them as authoritative sources of information. Although Ms. Phelan missed this segment of the conversation during the class, she viewed it while preparing her case study, and she was not concerned

Epistemological resources and framing424

about the students’ reliance on authority. She points to this fact in her case study, specifically with respect to vocabulary:

Here and later, it’s interesting to notice that they were still using big words, such as igneous, but now they were using those words in ways that made sense to them. (Phelan, 2006)

In this episode, the way in which students use the worksheets sup­ports rather than suppresses their sense­making (Rosenberg et al., 2006). Instead of focusing mostly on vocabulary and instead of sim­ply recording what the worksheets say, the students try to incorporate that information into the causal story they are telling. Specifically, when Johanna announces from the worksheet that a metamorphic rock arises from “heat and pressure,” Lisa and Bethany immediately won­der where the heat and pressure come from. Ryan provides an answer (“the Earth’s core”), and Johanna herself builds on Ryan’s answer (“it’s underground, so it’s closer to the core”). The sense­making discussion continues, and a minute later, when the students find themselves stuck, Bethany asks “what happens after that” and reads from the worksheet, “Immense heat and pressure, deep beneath the earth’s surface, change the rock into a metamorphic rock.” The argument about the source of that heat then reignites, with Johanna pointing out a possible contradic­tion in Ben and Gustavo’s account.

In summary, during both this conversation and the earlier one before Ms. Phelan’s intervention, students retrieve information from the work­sheets. In the latter conversation, however, the information feeds into the students’ collaborative knowledge generation, while in the earlier conversation, the information retrieval is the knowledge generation. This distinction illustrates the following point: when judging whether a behavior such as retrieving information from worksheets is productive for students’ learning, teachers (and researchers!) need to look not just at the behavior but at the overall pattern of activity.

The resources perspective invites and buttresses this conclusion. In the earlier conversation, students frame knowledge as bits of information to be retrieved from authority, an epistemological frame that undoubtedly includes knowledge as propagated stuff as a hub in the network of resources. During the latter conversation, the students’ epistemological frame still includes knowledge as propagated stuff; they continue to view the work­sheets as a source of knowledge that can be transmitted to them. But knowledge as propagated stuff no longer functions as a central, hub resource in the network. A dynamic model of this epistemological frame might have knowledge as propagated stuff becoming active only when the students cannot fill in the next causal link in the story they are constructing. In

Andrew Elby and David Hammer 425

this way, the idea of multiple epistemological frames drawing upon over-lapping sets of resources provides a framework for describing how trans­missionist ideas can be part of a productive constructivist stance.

Implications for instruction and research. This analysis has implications for both researchers and teachers trying to identify what counts as evi­dence of productive versus unproductive epistemologies.

First, probing students’ epistemologies as part of formative assess­ment, either through formal written instruments or informal obser­vations, a teacher should fight the general tendency to decompose the desired attitudes/beliefs/behaviors into pieces that are separately “tested.” Instead, the teacher should look more holistically at how stu­dents frame their activity.

Second, viewing frames as composed of finer grained resources allows analysis of the cognitive dynamics of epistemological “beliefs,” of what holds different frames together in different moments. For example, in the rock cycle episode, the teacher’s invocation to “start with what you know” pressed on one sort of resource, maybe including knowledge as constructed, leading to a locally stable epistemological frame that also afforded the retrieval of information from worksheets.

Third, epistemological frameworks that stand transmissionist views in stark opposition to constructivist views support overly simplistic characterizations of progressive instruction; teachers are exhorted “not to tell” (Chazan and Ball, 1999), so that students will figure things out for themselves. The framing perspective affords more nuance of inter­pretation and intervention. It can be perfectly appropriate for students to seek information from authority, and for teachers to provide it. The issue is how that retrieved information fits into the students’ overall framing of their epistemic activity. If the retrieved information feeds into a larger activity that the students frame as constructing a causal story, as illustrated previously, then seeking knowledge from authority was epistemologically productive.

The framing perspective provides guidance about fostering epistemological change

In this Section, we argue that the resources perspective provides guid­ance about how to foster epistemological change over both short and long timescales. These instructional implications stem from: (1) a view of frames as composed of finer grained epistemological resources; and (2) the three mechanisms of frame stability discussed previously.

Epistemological resources and framing426

According to our perspective, a teacher wishing to initiate the proc­ess of epistemological change might first try to unearth and activate productive sets of epistemological resources that the students already have but currently fail to activate consistently in that teacher’s class­room. Sometimes, a small epistemological “nudge” is enough to start the process. In Ms. Phelan’s class, the nudge was “start from what you know,” a brief intervention that was, evidently, sufficient to prompt the students’ reframing. Similarly, in our experience as physics teachers, we have found that telling a student to “Pretend you don’t know any physics” when answering a given question can (temporarily) make the student stop framing knowledge generation as retrieving facts and for­mulas from authority and start framing knowledge generation as figur­ing things out by sense­making (Hammer and Elby, 2003).

In general, a frame induced in this way will not be stable. To facilitate contextual stability of a productive frame, a teacher can tweak aspects of the classroom environment: providing opportunities for students to figure out concepts and problem­solving approaches for themselves, rewarding sense­making even when it leads to an incorrect answer, and otherwise creating a classroom culture that bolsters the productive epistemological frames the teacher is trying to induce (Hammer and Elby, 2003).

The teacher can reinforce this contextual stability by trying to induce deliberate stability, helping students to monitor their views of knowl­edge and learning as part of an attempt to maintain a productive fram­ing. For example, having students reflect upon their learning during the learning process, and not just at the end, can help them monitor their framing of the activity (Elby, 2001). At a deeper level, the teacher can involve students in the process of changing the classroom culture. For instance, in mathematics classes studied by Cobb and collaborators (Yackel and Cobb, 1996), the teacher and students jointly construct socio-mathematical norms, expectations about what counts as good math­ematical knowledge generation and argumentation. These norms have a strong epistemological orientation toward sense­making. As students begin to monitor their own adherence to these norms, they are among other things monitoring their epistemological framing of their class­room activity, which can help to stabilize a productive frame.

In the longer run, a frame repeatedly stabilized by contextual and deliberate means can become structurally stable, a compiled cogni­tive structure that turns on as a unit. For example, many students studied by Yackel and Cobb (1996) eventually learned to engage in mathematical sense­making and argumentation without needing con­stant reminders from themselves or from others about the classroom norms. For those students, the epistemological frame(s) associated

Andrew Elby and David Hammer 427

with the socio­mathematical norms may have become structurally sta­ble. Explicit reflection upon and meta­cognitive monitoring of their framing can perhaps speed up the development of structural stabil­ity. Notice that, by our account of this particular case, meta­cognition does not play the role of activating epistemological beliefs, as in Hofer (2004). Instead, meta­cognition serves to stabilize an already activated epistemological frame so that it can become a full­fledged epistemo­logical belief.

Structural stability, though, does not mean global applicability. It is possible that the same students who engage in deep sense­making in Yackel and Cobb’s math class “revert” to rote learning in their tra­ditionally taught classes, without being aware that they have shifted epistemologically. To develop a fully sophisticated epistemology, a stu­dent would need to have developed multiple structurally stable episte­mological frames and the ability to consciously manipulate and choose between them, according to context.

Notice that, in the resources perspective, naïve epistemological “beliefs” do not always need to be confronted and dismantled. Instead, the activation conditions of the frames corresponding to those stances could gradually evolve so that activation occurs – consciously, as described by Kuhn and Weinstock’s (2002) evaluativist stage, or uncon­sciously – only in appropriate contexts. For instance, when learning the state capitols, it is appropriate to view knowledge as stuff transmitted from authority and learning as accumulation of that stuff.

Are these resources and frames epistemological?

Given the above arguments, a critic could respond as follows:

Sure, Ms. Phelan’s intervention helped students reframe the activity, from recording information drawn from authority to constructing a causal story. But there’s no evidence that students’ epistemologies shaped their pre­ intervention or post­intervention behavior. It’s more likely that their pre­intervention behavior stemmed from habit, expediency (the worksheets were readily available), and/or expectations about what Ms. Phelan wanted. And it’s likely that their post­intervention behavior stemmed from their revised expectations and/or from a classroom routine that Ms. Phelan had previously established. They were just following her directions, not activating an epistemology.

We agree that habits and expectations play a role here and that students are following Ms. Phelan’s directions when they shift their approach. Our argument is that epistemology plays a role in their shift from one behavior to the other. Our evidence is that students are able to

Epistemological resources and framing428

understand and respond appropriately to the epistemological compo­nent of Ms. Phelan’s intervention. Part of her suggestion could be taken as referring purely to students’ behavior: stop looking at your papers and instead do something else. But the “something else” she suggests is not a specific behavior or activity; she does not instruct students to tell a story or explain the rock cycle in everyday language. She just suggests, “start with what you know, not with what the paper says.” For students to understand and act upon that suggestion in the way they do, not only starting with what they know but also using what they know to con­struct a causal story, they have to understand knowledge as something they have in their heads and as something that can be built upon, not just as information on the worksheets.

This insight about knowledge is not deep; we expect almost any eighth­grader to believe that they have knowledge in their heads and that in some cases they can build upon that knowledge. Nonetheless, this insight is epistemological in the sense that it is a view about knowl­edge. Furthermore, despite the seeming triviality of this epistemological insight, it was apparently dormant during the students’ pre­interven­tion discussion of Teutonic plates and the like. A purely non­episte­mological explanation of the students’ behavioral shift is implausible. The students do not “blindly” follow Ms. Phelan’s instructions; they continue to use the worksheets, as documented above. This counts as further evidence against the claim that students were simply following Ms. Phelan’s directions. The students could not have understood which habits/routines Ms. Phelan expected of them without understanding the epistemological insight that they have knowledge in their heads to be built upon.

Current and future research

From a resources perspective, it is essential to examine personal episte­mologies “not as a decontextualized set of beliefs, but as an activated, situated aspect of cognition that influences the knowledge construction process” (Hofer, 2004). We plan two general approaches to research on epistemological coherences and shifts: (1) naturalistic case studies; and (2) manipulations of context.

Naturalistic case studies

By this we mean observations of teachers teaching and students learn­ing in the classroom, sometimes supplemented by interviews. The study described above is one example. Another explores physics teachers’

Andrew Elby and David Hammer 429

epistemologies (Elby et al., in press; Lau et al., 2008). Lau videotapes classroom interactions, looking in particular for moments in which the teacher seems to shift from one stance to another. She chooses video snippets to watch and discuss with the teachers, prompting them to discuss what they might have been thinking at the time. The class­room and interview videos and transcript provide a corpus of data for a grounded theoretical approach: we identify epistemological categories in their statements and behavior and devise a scheme for coding them reliably. If a given teacher displays multiple epistemological coherences that “turn on” at different times during the class, based on plausible contextual factors, then the evidence favors a resources­based analysis of the teacher’s epistemology. By contrast, if a teacher’s epistemology remains consistent over a range of different classroom interactions and lesson types, except perhaps for a rare “blip,” the resources perspective adds nothing to a beliefs­ or stage­based account.

In another study based on a classroom video, Rachel Scherr is lead­ing analyses of college students working in collaborative groups on guided­inquiry worksheets in an algebra­based introductory physics course (Redish et al., 2003). Scherr discovered several distinct group “behavioral modes” indicated by gestures, posture, vocal pitch, and cadence, and other cues other than what the students were saying. With less than fifteen minutes of training, interrater reliability is 90 percent, to five second precision. Studies of four groups showed that student conversation about mechanism occurs disproportionately during a behavioral mode in which the students gesture prolifically, sit upright rather than slouched over the table, and look at each other rather than at the worksheet. (Informal analysis of the rock cycle video suggests a different set of behavioral modes but a similar interaction of substance and gestures.) A similar study will independently code for occurrences of that behavioral mode and for a sense-making epistemological framing, in which students treat physics knowledge as something coherent that they can figure out. Such studies can shed light on behavioral indica­tors for certain frames.

Manipulations of context

We have also begun to study epistemological coherences and shifts in contexts we design for research. Lising and Elby (2005) analyzed the reasoning of one student, “Jan,” whose work in that same introduc­tory physics class (Redish et al., 2003) gave evidence that she framed classroom physics knowledge as disconnected from everyday intuitive reasoning. They then created a clinical setting – a comfortable room

Epistemological resources and framing430

in the education building rather than a classroom in the physics build­ing, with the interviewer introduced as an education researcher rather than a physicist – to study the effects of context on Jan’s epistemo­logical framing. Working on similar physics problems in this different setting, Jan had no trouble making use of everyday intuitive reason­ing. The clinical interview induced a different sort of epistemic activ­ity than she displayed in physics class. Case studies of how individual students’ epistemologies change or remain stable across different con­texts have implications for teachers trying to nudge students’ episte­mologies and researchers trying to build cognitive accounts of personal epistemologies.

We are also developing large­N manipulations. One of our students, Paul Hutchison, devised a plan based on a question we have used in the classroom:

Two identical balls are thrown from the same height at the same speed. One ball is thrown sideways (horizontally) while the other is thrown straight down. Which ball, if either, lands first?

Of course, the ball thrown downward reaches the ground first. How­ever, when we give this question in the first lecture of a large, second­semester physics course (N = 120–80), as many as half the students say the balls land at the same time, invoking (incorrect) equation­based reasoning or an incorrectly remembered fact from a previous phys­ics class. When we frame the question differently, asking students to “Pretend you’ve never studied physics,” essentially everyone gives the correct answer, relying on common­sense reasoning. By interviewing students, having them think aloud while solving the problem and ask­ing follow­up questions such as “How did you get your answer?” and “How do you know that’s right?” we could gather evidence bearing on our hunch that students’ two distinct approaches to the problem reflect two different epistemological frames, a sense­making frame involving knowledge as constructed versus an “answer­making” frame centered around knowledge as propagated stuff.

Hutchison’s idea was to devise two questionnaires that might tip students into one epistemological frame or the other. He and another student, Renee Michelle Goertzen, developed pilot versions, each including three questions. On questionnaire A, the first two ques­tions were designed to invoke “physics class,” while on questionnaire B the first two questions tried to invoke “interesting puzzle.” On both questionnaires, the third (target) question was identical, the problem presented above about the two thrown balls. Their pilot study gave a signal: students working from the “puzzle” version were much more

Andrew Elby and David Hammer 431

likely to give the sensible answer on the thrown balls question. We have experimented with different prompts, with mixed results, but this is a direction we hope to pursue. In this work we are inspired by Claude Steele and his colleagues’ abilities to manipulate students’ experience of stereotype threat and show its influence on their performance (Sackett et al., 2004; Steele, 1997).

Conclusion

We argued that the resources perspective on personal epistemology is particularly productive for teachers to learn, for three reasons. First, epistemological frames correspond to what teachers can recognize, and novice teachers can learn to recognize, in students’ approaches to learning. Second, the possibility that a given resource can participate in multiple frames invites close attention to context when evaluating whether a given student utterance or behavior reflects a productive stance toward knowledge, leading to more nuanced assessment of the student’s approach to learning. Third, the resources perspective pro­vides guidance about fostering epistemological change over both short and long time scales.

Although the general warrant for “knowledge in pieces” frameworks stems from cognitive theory and clinical experimentation (diSessa, 1993; Karmiloff­Smith, 1992), many of the emerging details of the resources perspective as applied to personal epistemologies come from min­ing the knowledge and actions of teachers (Hammer and Elby, 2003). This theory­building design principle stems from our conviction that, when it comes to understanding students’ approaches to knowledge and learning, the (tacit) practitioner knowledge underlying produc­tive instructional moves such as Ms. Phelan’s rock cycle intervention is often “ahead” of researchers’ theories. Just as teachers learning about resources and framing can unearth, build upon, and refine their practi­tioner knowledge, we as theorists unearth, build upon, and refine teach­ers’ practitioner knowledge in generating our theoretical perspective.

R EF ER ENCES

Bendixen, L. D. and Rule, D. C. (2004). An integrative approach to personal epistemology: A guiding model. Educational Psychologist, 39(1), 69–80.

Carey, S. (1992). The origin and evolution of everyday concepts. In R. N. Giere (Ed.), Cognitive Models of Science (Vol. XV, 89–128). Minneapolis: Univer­sity of Minnesota Press.

Chazan, D. and Ball, D. L. (1999). Beyond being told not to tell. For the Learn-ing of Mathematics, 19(2), 2–10.

Epistemological resources and framing432

Corno, L. (1993). The best­laid plans: Modern conceptions of volition and educational research. Educational Researcher, 22(2), 14–22.

diSessa, A. A. (1993). Towards an epistemology of physics. Cognition and Instruction, 10(2–3), 105–225.

diSessa, A. A., Elby, A., and Hammer, D. M. (2002). J’s epistemological stance and strategies. In G. M. Sinatra and P. R. Pintrich (Eds.), Intentional con-ceptual change (237–290). Mahwah, NJ: Lawrence Erlbaum Associates.

Elby, A. (2001). Helping physics students learn how to learn. American Journal of Physics, 69(7 SUPP1), S54–64.

(2009). Defining personal epistemology: A response to Hofer and Pintrich (1997) and Sandoval (2005). Journal of Learning Sciences, 18(1), 139–49.

Elby, A. and Hammer, D. (2001). On the substance of a sophisticated episte­mology. Science Education, 85(5), 554–67.

Elby, A., Lau, M., and Hammer, D. M. (submitted). Accounting for variability in a teacher’s epistemology.

Hammer, D. M. (1994). Epistemological beliefs in introductory physics. Cogni-tion and Instruction, 12(2), 151–83.

(2000). Student resources for learning introductory physics. American Jour-nal of Physics, Physics Education Research Supplement, 68(S1), S52–9.

Hammer, D. M. and Elby, A. (2002). On the form of a personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychol-ogy of beliefs about knowledge and knowing (169–90). Mahwah, NJ: Law­rence Erlbaum Associates.

(2003). Tapping epistemological resources for learning physics. Journal of the Learning Sciences, 12(1), 53–90.

Hammer, D. M., Elby, A., Scherr, R. E., and Redish, E. F. (2005). Resources, framing, and transfer. In J. Mestre (Ed.), Transfer of learning from a modern multidisciplinary perspective (89–120). Greenwich, CT: Information Age Publishing.

Hewson, P. W. (1981). A conceptual change approach to learning science. Euro-pean Journal of Science Education, 3(4), 383–96.

Hofer, B. K. (2004). Epistemological understanding as a metacognitive proc­ess: Thinking aloud during Online searching. Educational Psychologist, 39(1), 43–55.

Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learn­ing. Review of Educational Research, 67(1), 88–140.

Hogan, K. (1999). Relating students’ personal frameworks for science learn­ing to their cognition in collaborative contexts. Science Education, 83(1), 1–32.

Karmiloff­Smith, A. (1992). Beyond Modularity. Cambridge, MA: MIT Press.

King, P. M. and Kitchener, K. S. (2002). The reflective judgment model: Twenty years of research on epistemic cognition. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (37–61). Mahwah, NJ: Lawrence Erlbaum Associates.

Andrew Elby and David Hammer 433

Kuhn, D. (1991). The skills of argument. Cambridge: Cambridge University Press.

Kuhn, D. and Weinstock, M. (2002). What is epistemological thinking and why does it matter? In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing (121–44). Mahwah, NJ: Lawrence Erlbaum Associates.

Lau, M., Elby, A., and Hammer, D. M. (2008). Attending to students’ ideas: Two case studies of how teachers attend to student ideas. Paper pre­sented at the International Conference of the Learning Sciences, Utrecht, Netherlands.

Lising, L. and Elby, A. (2005). The impact of epistemology on learning: A case study in introductory physics. American Journal of Physics, 73(4), 372–82.

Louca, L., Elby, A., Hammer, D. M., and Kagey, T. (2004). Epistemological resources: Applying a new epistemological framework to science instruc­tion. Educational Psychologist, 39(1), 57–68.

May, D. B. and Etkina, E. (2002). College physics students’ epistemological self­reflection and its relationship to conceptual learning. American Jour-nal of Physics, 70(12), 1249–58.

McCloskey, M. (1983). Naive theories of motion. In D. Gentner and A. Stevens (Eds.), Mental models (299–324). Hillsdale, NJ: Lawrence Erl­baum Associates.

Muis, K. R., Bendixen, L. D., and Haerle, F. C. (2006). Domain­generality and domain­specificity in personal epistemology research: Philosophical and empirical reflections in the development of a theoretical framework. Educational Psychology Review, 18(1), 3–54.

Phelan, J. (2006). Eighth graders discuss the rock cycle. In D. M. Hammer and E. H. van Zee (Eds.), Seeing the science in children’s thinking: Case studies of inquiry in physical science (96–115). Portsmouth, NH: Heinemann.

Redish, E. F. (2004). A theoretical framework for physics education research: Modeling student thinking. In E. F. Redish et al. (Eds.), Proceed-ings of the Enrico Fermi Summer School, Course CLVI (1–63). Bologna: Ital­ian Physical Society.

Redish, E. F., McCaskey, T., Hammer, D., Elby, A., and Lising, L. (2003). Epis­temological gains in a large lecture class. Paper presented at the American Association of Physics Teachers National Meeting, Madison, WI.

Rosenberg, S. A., Hammer, D., and Phelan, J. (2006). Multiple epistemologi­cal coherences in an eighth­grade discussion of the rock cycle. Journal of the Learning Sciences, 15(2), 261–92.

Sackett, P. R., Hardison, C. M., and Cullen, M. J. (2004). On interpreting ster­eotype threat as accounting for African American–white differences on cognitive tests. American Psychologist, 59(1), 7–13.

Sandoval, W. A. (2005). Understanding students’ practical epistemologies and their influence on learning through inquiry. Science Education, 89(4), 634–56.

Schommer, M. (1993). Epistemological development and academic perform­ance among secondary students. Journal of Educational Psychology, 85(3), 406–11.

Epistemological resources and framing434

Schommer, M., Crouse, A., and Rhodes, N. (1992). Epistemological beliefs and mathematical text comprehension: Believing it is simple does not make it so. Journal of Educational Psychology, 84, 435–43.

Steele, C. M. (1997). A threat in the air: How stereotypes shape intellectual identity and performance, American Psychologist, 52(6), 613–29.

Tannen, D. (1993). Framing in discourse. New York: Oxford University Press.Thelen, E. and Smith, L. B. (1994). A dynamic systems approach to the develop-

ment of cognition and action. Cambridge, MA: MIT Press.Yackel, E. and Cobb, P. (1996). Sociomathematical norms, argumentation,

and autonomy in mathematics. Journal for Research in Mathematics Educa-tion, 27, 458–77.

435

14 The effects of teachers’ beliefs on elementary students’ beliefs, motivation, and achievement in mathematics

Krista R. Muis and Michael J. Foy McGill University

Introduction

According to Pajares (1992), teachers’ beliefs can be deeply personal, unaffected by persuasion, and either implicitly or explicitly expressed in daily routines. Their beliefs can be formed by chance, an intense expe­rience or a succession of events, and may include beliefs about different facets of teaching and learning. Teachers hold beliefs about students, learning, teachers and teaching, the nature of knowledge and know­ing, the roles of schools in society, and the curriculum, to name a few (Levitt, 2001). Whatever their origin or object, research has shown that beliefs influence a wide variety of cognitive processes including mem­ory, comprehension, deduction and induction, problem representation, and problem solution (Pintrich, 1990). Importantly, the study of teach­ers’ beliefs provides a valuable means of analyzing and understanding the complex relationship between beliefs and student outcomes (Hofer and Pintrich, 2002; Pajares, 1992; Schraw and Olafson, 2002).

In his review of research on teachers’ beliefs, Pajares (1992) reported that teachers’ beliefs about teaching and learning, including beliefs about students, significantly influence teachers’ classroom practices. Moreover, he found that teachers’ beliefs are more likely to influence the types of instructional strategies they implement in the classroom than their knowledge about a particular content area or instructional strategies. As Peterman (1993) and Tobin (1993) observed, the primary way in which teachers’ educational beliefs are given meaning is through their expression in the classroom.

Historically, definitions of teachers’ beliefs have included a number of constructs (e.g., dispositions, perspectives, opinions, judgments, and knowledge) but over time the focus has shifted to emphasize the object of the belief. For example, specific definitions include beliefs about one’s

Effects of teachers’ beliefs on elementary students436

confidence to affect performance – efficacy beliefs; beliefs about the nature of knowledge and knowing – epistemic beliefs; beliefs about teaching and learning; and beliefs about particular content areas like reading, science, and mathematics (Pajares, 1992). It is important to note that researchers have labeled teachers’ personal beliefs about knowledge and knowing as epistemic beliefs, but have similarly labeled teachers’ beliefs about how students acquire knowledge as epistemic beliefs about teaching and learn­ing. That is, some researchers have measured teachers’ domain­general or domain­specific epistemic beliefs (e.g., Tsai, 2006); teachers’ own personal beliefs about knowledge and knowing. In contrast, others have measured teachers’ beliefs about how students acquire knowledge. Typically, when the focus is on teachers’ beliefs about how students acquire knowledge, researchers measure whether teachers believe that students can actively construct knowledge or whether students are passive recipients of knowl­edge (e.g., Hashweh, 1996). Moreover, in some studies (e.g., Buehl et al., 2002; Schraw and Olafson, 2002), instruments used to measure teachers’ beliefs include both teachers’ personal views about knowledge and know­ing as well as their beliefs about how students acquire knowledge. For the purposes of this paper, we refer to constructivist epistemic beliefs as those that reflect an individual’s beliefs in the tentative nature of knowledge, the complexity of knowledge, and the active construction of knowledge. Moreover, we refer to teachers’ beliefs about teaching and learning as their beliefs about how students acquire knowledge. In particular, constructiv­ist beliefs about teaching and learning refer to those that reflect a belief that students actively construct knowledge.

Regardless of the focus, these beliefs are important to consider in the broader classroom context. More specifically, several studies have examined the relationship between epistemic beliefs and beliefs about teaching and learning, and have assessed how these beliefs influence practice. Within this line of work, researchers have found support for the notion that teachers’ epistemic beliefs influence their beliefs about teaching and learning as well as their instructional approaches (e.g., Brickhouse, 1989; Hammrich, 1998; Hashweh, 1996; Lederman, 1992; Linder, 1992; Tsai, 2002, 2006). For example, Hashweh (1996) meas­ured teachers’ beliefs about students’ learning and conceptual change in science and found that teachers who held more constructivist beliefs (e.g., students are capable of actively constructing their own knowl­edge) were more likely to explore students’ alternative conceptions of scientific phenomena, had a richer repertoire of instructional strategies, and were more likely to use instructional strategies to foster conceptual change than teachers with more traditional views (e.g., students are passive recipients of knowledge, which is handed down by an authority

Krista R. Muis and Michael J. Foy 437

figure). Similarly, Daniels and Shumow (2003) reported that teachers’ epistemic beliefs influence their practice and the learning environment within which students learn, and that the environment influences stu­dents’ behaviors. Based on their review of the literature, they concluded that students in constructivist­oriented environments (student­focused) have increased motivation, decreased stress, and increased problem­solving and language skills compared to students in traditional (e.g., teacher­focused) learning environments.

Although a number of studies have proposed direct links among teachers’ epistemic beliefs, teaching and learning beliefs, and their instructional practices, not all studies have provided empirical sup­port (e.g., Benson, 1989; Duschl and Wright, 1989; Lederman, 1999; Lederman and Zeidler, 1987; Levitt, 2001; Schraw and Olafson, 2002; White, 2000; Wilcox­Herzog, 2002). These studies reported several constraints that affected the coherence between beliefs and practice such as social factors (Duschl and Wright, 1989), situational constraints (Benson, 1989), teachers’ level of experience, intentions, and perceptions of students (Lederman, 1999), and a reliance on district­mandated curriculum and teaching strategies (Schraw and Olafson, 2002).

The broader educational implications of these studies are important. Across all content areas, calls have been made for teachers to imple­ment more constructivist­oriented instructional techniques to improve student learning (e.g., National Council of Teachers of Mathematics (NCTM), 2006; National Science Board, 2006). If teachers’ beliefs influence the types of instructional strategies they use, then interven­tions are needed to help foster epistemological development in teachers toward more constructivist beliefs, which may improve student learn­ing. Of particular concern, as Muis (2004) identified in her review of students’ beliefs about mathematics knowledge, students at all levels of education hold non­constructivist beliefs about mathematics knowl­edge. For example, when asked about the certainty of mathematics knowledge, students believe knowledge is unchanging. Students also believe mathematics knowledge is passively handed to them by some authority figure, typically the teacher or textbook (Garofalo, 1989), and believe they are incapable of learning mathematics through logic or reason. Another common belief is that various components of math­ematics knowledge are unrelated; the structure consists of isolated bits and pieces of information (Garofalo, 1989). Students do not believe they are capable of constructing mathematical knowledge and solv­ing problems on their own (Schoenfeld, 1988). Finally, students typi­cally believe learning mathematics should be quick, within five to ten

Effects of teachers’ beliefs on elementary students438

minutes (Frank, 1988). If they have not solved the problem or come up with the correct answer in that time period, students give up.

To determine why students espouse these non­constructivist epis­temic beliefs about mathematics, Schoenfeld (1988) examined the classroom environment within which students learn mathematics. He conducted a year­long intensive study in a suburban school district to examine the presence and robustness of four beliefs about mathemat­ics typically found among students and to seek the possible origins of those beliefs. Based on his analysis of classroom processes, observa­tions, and interviews, Schoenfeld concluded that instructional strat­egies used in the classroom had a profound influence on students’ epistemic beliefs.

Indeed, several studies have examined how instructional strategies (e.g., Doyle, 1988; Franke and Carey, 1997; Schoenfeld, 1988; Stodol­sky, 1985) and changes in those strategies (e.g., Carter and Yackel, 1989; Higgins, 1997; Hofer, 1999; Lampert, 1990) influence students’ epistemic beliefs about mathematics. For example, Hofer (1999) exam­ined relations among students’ beliefs and motivation, learning strate­gies, and academic performance in two different instructional contexts in introductory calculus. One instructional context used traditional methods; instructors used a standard calculus text that proceeded sequentially, and were expected to cover a required amount of material primarily through lectures and demonstrations of problem sets. The alternative instructional context, called the “new wave” approach, used more social­constructivist approaches where collaborative learning was emphasized, students engaged in active learning and were expected to work situated problems with potentially multiple approaches and more complex solutions. Hofer found that, for students, more constructivist epistemic beliefs were positively correlated with intrinsic motivation, self­efficacy, and self­regulation, as well as with course grades. More­over, at the end of the term students enrolled in the “new wave” sections exhibited more constructivist epistemic beliefs than students enrolled in the traditional style instruction sections.

In our review of the literature, we found only two studies that exam­ined direct relations between teachers’ epistemic beliefs and students’ epistemic beliefs (e.g., Johnston et al., 2001; Tsai, 2006). For exam­ple, Tsai (2006) interviewed junior high school teachers to identify their epistemic beliefs about science, conducted classroom observa­tions during science instruction, and had students complete self­ reports designed to measure their epistemic beliefs about science and their perceptions toward the science learning environment. Based on

Krista R. Muis and Michael J. Foy 439

qualitative and quantitative analyses, Tsai found coherences between teachers’ epistemic beliefs about science and their instructional prac­tices. Teachers who were more positivist­oriented in their beliefs about science (e.g., science knowledge is “discovered,” science knowledge is certain) focused on students’ test scores, spent more time on teacher­directed lectures, tutorial problem practices, or in­class examinations. In contrast, teachers with more constructivist epistemic beliefs about science knowledge (e.g., science knowledge is tentative, scientists create knowledge, social negotiation is important to the justification process) focused more on student application and understanding of content, and spent more time on student­focused inquiry or interactive discussion. Finally, Tsai reported that some evidence was found that teachers’ epis­temic beliefs might be related to students’ epistemic beliefs, but that coherence between the two was not obvious.

Although Tsai (2006) examined relations between teachers’ epis­temic beliefs, instructional practices, and students’ epistemic beliefs, the analytic techniques used indirectly measured coherence between teachers’ and students’ beliefs. As Pintrich (2002) noted, more studies are needed that include advanced statistical techniques, such as struc­tural equation modeling, to make stronger causal claims than those afforded with more qualitative or indirect techniques. Muis (2004) and Muis et al. (2006) also suggest that to improve understanding of the nature of epistemic beliefs, more research is needed that examines how teachers’ beliefs might directly influence students’ beliefs and other facets of student learning like motivation and achievement. Moreover, as Burr and Hofer (2002) highlighted, researchers who study epistemic beliefs have been criticized for the lack of research with younger partici­pants. Accordingly, the purpose of our study was to address these gaps in the literature. We examined relations between teachers’ epistemic and teaching and learning beliefs about mathematics, and elemen­tary students’ achievement goal orientations, self­efficacy, epistemic and learning beliefs, and achievement in the context of mathematics problem­solving.

To hypothesize how teachers’ beliefs might influence students’ beliefs, motivation, and achievement, we used Biggs’ (1993) “3P” sys­tems model (e.g., presage, process, product) of teaching and learning, and Schommer­Aikins’ (2004) embedded systemic model of epistemic beliefs. We briefly describe each model, discuss how the facets we examined may be related, and present empirical evidence that supports these hypothesized relations. The section ends with our research ques­tions and hypotheses.

Effects of teachers’ beliefs on elementary students440

Biggs’ 3P model

According to Biggs (1993), education is a system of interacting compo­nents, which include the students, teachers, teaching contexts, student learning processes, learning outcomes, administrators, and any other component that affects learning. In his model, Biggs describes a lin­ear progression from presage to process to product (i.e., 3P) whereby each component interacts with all other components to form a system. Student presage factors are relatively stable, learning­related charac­teristics that include, for example, prior knowledge, ability beliefs, and motivational beliefs. Teacher presage factors are contextual, and include teacher personality, classroom climate, course structure, curriculum content, and instructional strategies. Accordingly, we posit that stu­dents’ and teachers’ epistemic and learning beliefs are one component of the presage factors in Biggs’ model.

The next component in Biggs’ (1993) model is the process compo­nent. Processes include cognitive mechanisms used to learn such as the types of learning approaches students engage in to complete a task. Biggs identified three approaches: surface, deep, and achieving. A deep approach to learning is task­centered and task­appropriate, and includes strategies such as elaboration of information. The sur­face approach is based on a motive to minimize effort and to minimize the consequences of low effort; strategies include rote memorization of information. Finally, the third process, an achieving approach, is outcome­oriented; good grades are the goal and students who adopt this approach are most concerned about the cost–effective use of their time and effort. All three types of strategies students use to learn subse­quently affect performance – one type of outcome or product, the final component of Biggs model.

Schommer-Aikins’ embedded systemic model

In 1990, Schommer developed a framework to take into consideration the potential multifaceted nature of epistemic beliefs. She proposed that beliefs about knowledge and learning include multiple dimensions, with each dimension being relatively independent of the others. She hypoth­esized three dimensions of beliefs about knowledge that span along continua on: (1) the certainty of knowledge, ranging from knowledge is unchanging to knowledge is evolving; (2) the source of knowledge, ranging from knowledge is handed down by authority to knowledge is acquired through reason or logic; and (3) the simplicity of knowl­edge, ranging from knowledge is organized as isolated bits and pieces to

Krista R. Muis and Michael J. Foy 441

knowledge is organized as highly interrelated concepts. She proposed two further dimensions related to learning. The fourth dimension of her model is (4) the control of knowledge acquisition, ranging from the ability to learn is inherited and unchangeable to the ability to learn can improve over time. The last dimension is (5) the speed of knowledge acquisition, ranging from learning is quick or not at all to learning is gradual.

Since 1990, Schommer has continued to develop her framework, and recently presented a more complex model to account for factors that may influence and nature and development of epistemic beliefs. Simi­lar to Biggs (1993), Schommer­Aikins (2004) proposed that a number of social factors, such as teachers, peers, and family, have an impor­tant influence on student learning. To better understand the nature of epistemic beliefs, she highlighted the need to take into considera­tion various social factors, cultural views (e.g., collectivist versus indi­vidualist views), as well as other cognitive and affective factors. More specifically, Schommer­Aikins stressed the importance of assessing the influences of exogenous variables, like culture, as they can directly and indirectly affect beliefs about ways of knowing, epistemic beliefs, class­room performance, and self­regulated learning. In her view, to consider epistemic beliefs in isolation does not provide an accurate account of their influence on learning, given that epistemic beliefs do not function in a vacuum.

Accordingly, based on Schommer­Aikins’ (2004) model, as well as Biggs’ (1993) model, multiple factors should be taken into considera­tion to examine what influences students’ epistemic beliefs, and how those influences interact to affect other educational outcomes such as performance. Specifically, both theorists propose that each com­ponent of their model interacts with and affects other components of the model, either directly or indirectly. We posit that teachers’ beliefs directly influence students’ beliefs, which subsequently influence stu­dents’ motivation and achievement. We further predict, consistent with Biggs’ and Schommer­Aikins models, that teachers’ beliefs will also directly influence students’ academic performance.

Epistemic beliefs, achievement goal orientations, and self-efficacy

We chose Buehl et al.’s (2002) theoretical framework to explore rela­tions between epistemic beliefs, motivation, and achievement. Buehl et al. developed a theoretical framework that extends Schommer’s (1990) original work, but that takes into consideration the domain­specificity

Effects of teachers’ beliefs on elementary students442

of epistemic beliefs. As Buehl et al. illustrate, a number of research­ers have demonstrated that contextual factors, including the academic domain, have an important influence on learning outcomes (e.g., Lam­bert and McCombs, 1998). Accordingly, they constructed a domain­specific epistemic beliefs questionnaire that focuses solely on academic epistemic beliefs. Two broadly articulated domains were chosen, math­ematics and history, on which to focus their instrument.

Based on a series of studies Buehl et al. (2002) conducted to exam­ine the validity and reliability of their instrument, they identified two dimensions within each domain, with a total of four: two that focused on beliefs about knowledge acquisition (one for mathematics, and one for history), and two that focused on the integration of information (again, one for each domain). Similar to Schommer’s (1990) “control of knowledge acquisition” and “simplicity of knowledge” dimensions, they labeled their dimensions “need for effort” and “integration of infor­mation and problem­solving,” respectively, for each domain. We chose their framework given the domain­specific nature of the items, but also chose this framework given that items focus both on teachers’ personal beliefs about knowledge and knowing as well as beliefs about how stu­dents learn and what strategies are helpful to improve learning.

Using Buehl et al.’s (2002) instrument, we sought to explore relations between teachers’ beliefs about mathematics, students’ beliefs about mathematics, students’ achievement goals, their self­efficacy for prob­lem­solving in mathematics, and their mathematics achievement. To hypothesize relations, we drew on a number of theorists’ empirical and theoretical work. For example, Hofer and Pintrich (1997), Buehl (2003), and Schommer (1998; Schommer­Aikins et al., 2005) have suggested that learners’ epistemic beliefs may influence students’ motivation for learn­ing. More specifically, Hofer and Pintrich (1997) proposed that epis­temic beliefs may function as implicit theories that can induce particular types of goals for learning, such as mastery or performance approach oriented goals. According to Dweck and Leggett (1988), learners with a mastery goal orientation are theorized to believe ability is incremental; more specifically, ability is malleable, and effort can increase ability and the odds of succeeding in tasks. In contrast, a performance approach goal orientation theoretically is associated with an entity theory of abil­ity; that is, ability is fixed at a given level, this innate ability is necessary for success, and effort does not improve performance. The focus for mastery­oriented learners is to understand the content. In contrast, the focus for performance approach oriented individuals is performing bet­ter than others. Thus, if individuals believe the ability to learn is inher­ited and unchangeable, then they should adopt a performance approach

Krista R. Muis and Michael J. Foy 443

goal orientation. In contrast, if individuals believe that ability is malle­able, then they should adopt a mastery goal orientation.

Buehl (2003) further hypothesized that not only do epistemic beliefs influence goal orientations, but also task value beliefs and self­efficacy beliefs either directly or indirectly. Self­efficacy is the belief an indi­vidual has about his or her ability to organize and execute a course of action required to produce a desired outcome (Bandura, 1997). Theo­retically, a positive relationship between a mastery goal orientation and self­efficacy can be hypothesized to result from mastery­oriented indi­viduals’ tendency to believe ability is incremental (Button et al., 1996) and effort is a primary cause of success (Ames and Archer, 1988). In contrast, those with a performance approach orientation believe abil­ity is fixed. These individuals may doubt that past success will lead to future success on a particular task. Moreover, when negative feedback is given on one task and a subsequent task is perceived to be equally challenging, performance­oriented individuals may interpret negative feedback to imply poor future performance. This may arouse negative affect, which may negatively influence future performance.

Empirical studies support these hypotheses (e.g., Bråten and Olaus­sen, 2005; Buehl, 2003; Hofer, 1999; Neber and Schommer­Aikins, 2002). For example, Neber and Schommer­Aikins (2002) examined relations between epistemic beliefs and self­regulated science learning with a sample of highly gifted elementary and high school students. Self­reports were completed, by 133 gifted students, designed to meas­ure their regulatory strategy use, motivation, epistemological intentions, epistemic beliefs, and perceptions of their learning environment in sci­ence (for elementary students) and physics (for high school students).

Path analyses revealed that the extent to which students engaged in exploration in their science­learning environment (i.e., investigation) was a strong and direct predictor of regulatory strategy use. Episte­mological intention for facts was also a strong and direct predictor of regulatory strategy use. Moreover, investigation had an indirect effect on regulatory strategy use via the epistemic belief that success does not require work and via self­efficacy for learning science. Specifically, the more students experienced investigation in science, the more they believed that success requires effort. The more they believed that suc­cess requires effort, the greater their self­efficacy for learning science. Subsequently, the higher their self­efficacy, the greater their regulatory strategy use. Finally, the more individuals experienced investigation, the higher their mastery goal orientation, and the higher their mas­tery goal orientation, the higher their self­efficacy for learning science. Neber and Schommer­Aikins (2002) concluded that exploration and

Effects of teachers’ beliefs on elementary students444

discovery should be enabled and strengthened in science classrooms, particularly with gifted students.

The current study

Previous research has demonstrated that teachers’ beliefs influence the types of instructional strategies they use (Hashweh, 1996; Linder, 1992; Tsai, 2002, 2006). Research has also shown that the types of instruc­tional strategies teachers use influence students’ epistemic beliefs (e.g., Hofer, 1999; Schoenfeld, 1988; Stodolsky, 1985). To date, only two studies have examined the relationship between teachers’ and students’ epistemic beliefs (Johnston et al., 2001; Tsai, 2006). As Muis et al. (2006) and Bendixen and Rule (2004) suggest, teachers’ epistemic and learning beliefs are an important component of the epistemic climate of a classroom. To improve understanding of how the classroom epistemic climate can influence various facets of motivation and achievement, we extend Johnston et al.’s (2001) and Tsai’s (2006) work by examining relations between teachers’ epistemic and teaching and learning beliefs and elementary students’ achievement goal orientations, self­efficacy, epistemic and learning beliefs, and achievement in the context of math­ematics problem­solving.

Elementary school teachers completed Buehl et al.’s (2002) ques­tionnaire to measure their beliefs about the need for effort to learn mathematics and the integration of information and problem­solving in mathematics. Students in grades four and five completed a fifteen­item self­report questionnaire designed to measure beliefs about the need for effort to learn mathematics, the certainty and simplicity of mathematics knowledge, and their mastery and performance approach goal orientations. Students also completed a task­specific self­efficacy questionnaire for completing mathematics problems. Finally, students completed a fifteen­item mathematics test; items were drawn from standardized achievement tests for grades four and five.

Three research questions were examined in this study: are teach­ers’ epistemic and teaching and learning beliefs related to students’ epistemic and learning beliefs and student achievement? Do students’ epistemic and learning beliefs predict students’ levels of achievement goals? Does self­efficacy mediate the relationship between achievement goals and achievement? Given the theoretical frameworks and empiri­cal results we reviewed, we predicted a positive relationship between teachers’ beliefs about the need for effort to learn mathematics and students’ beliefs about the need for effort to learn mathematics, and a negative relationship between teachers’ beliefs about the integration of

Krista R. Muis and Michael J. Foy 445

information in mathematics and students’ beliefs in the certainty and simplicity of mathematics knowledge. Consistent with Biggs’ (1993) and Schommer­Aikins’ (2004) models, we also predicted that teachers’ beliefs would positively predict student performance in mathematics. We further predicted that the more students believe in the need for effort to learn mathematics, the less likely they are to adopt a perform­ance approach goal orientation. Moreover, we hypothesized that self­efficacy will mediate the negative relationship between a performance approach goal orientation and performance. In contrast, we expected a positive relationship between students’ beliefs about the need for effort to learn mathematics and a mastery goal orientation, and a posi­tive mediating relationship between a mastery goal orientation, self­efficacy, and achievement. Finally, we expected a negative relationship between students’ beliefs in the certainty and simplicity of mathematics knowledge and a mastery goal orientation, and a positive relationship between certain and simple knowledge and a performance approach goal orientation. The hypothesized model is presented in Figure 14.1.

Methodology

Participants

Participant teachers. Fifty­five grade four and five elementary school teachers from twenty­six different schools from a southwestern region

Teachers’belief ineffort

Students’belief ineffort

Teachers’belief inintegration

Students’belief incertain/simpleknowledge

Masterygoals

Performancegoals

Self-efficacy Achievement

+

–

–

–

+

+

+

–

+

+

+

Figure 14.1: Hypothesized model

Effects of teachers’ beliefs on elementary students446

of the United States volunteered to participate. All schools were iden­tified as Title I schools for the 2004–5 academic school year. Title I schools are those that have a large enrollment of students from a low socio­economic status (e.g., 60 percent or higher). Of these teach­ers, forty­eight were female and seven were male. Teachers’ mean age was 38.84 years (SD = 10.50) and the average amount of time teach­ing at their current schools was 2.81 years (SD = 2.65). The average total time teaching was 8.23 years (SD = 8.53). Of the total sample of fifty­five teachers, fourteen were currently teaching a combined grade four–five class.

Participant students. From the same teachers’ classrooms described above, 131 elementary school students from grades four and five volun­teered to participate. Teachers chose two to three students from their class to participate. Of these students, eighty­three were in grade four (forty­five females, thirty­eight males) and forty­eight were in grade five (twenty­eight females, twenty males), with forty­one of the 131 stu­dents in a combined grade four–five class. The average age of the grade four students was 9.54 years (SD = .63) and the average age of the grade five students was 10.58 years (SD = .58).

Materials

Teachers’ beliefs: the domain-specific belief questionnaire. To measure teach­ers’ beliefs about mathematics, Buehl et al.’s domain­specific belief questionnaire (DSBQ, 2002) was used. The DSBQ is a twenty­two­item self­report instrument designed to measure individuals’ beliefs about the need for effort in learning mathematics (e.g., constructivist epistemic view, five items; a construct similar to Schommer’s (1990) control of knowledge acquisition dimension), the integration of infor­mation and problem­solving in mathematics (e.g., constructivist epis­temic view, six items; similar to Schommer’s simplicity of knowledge dimension), the need for effort in learning history (five items), and the integration of information and problem­solving in history (six items). Individuals rate each item on a seven­point rating scale ranging from “strongly disagree” (a rating of one) to “strongly agree” (a rating of seven). Because mathematics was the focus of this study, participant teachers responded only to those items that referred to mathematics. A sample item from the integration of information subscale is “There are links between mathematics and other disciplines.” A sample item from the need for effort subscale is “Students who are good at math have to work hard.” (See Buehl et al. (2002) for a complete list of the items.)

Krista R. Muis and Michael J. Foy 447

Research that has assessed the reliability and validity of the DSBQ has found support for the proposed dimensions, good reliability esti­mates, and construct and predictive validity (e.g., Buehl et al., 2002). For example, internal consistency coefficients for the two mathematics subscales were .70 for the integration subscale and .68 for the need for effort subscale in mathematics.

Students’ epistemic beliefs, achievement goals, and self-efficacy. Students completed a fifteen­item questionnaire designed to measure their beliefs about the need for effort to learn mathematics (e.g., constructivist epis­temic view, four items), the certainty and simplicity of knowledge in mathematics (non­constructivist epistemic view, five items), and their mastery (three items) and performance approach goal orientations (three items) in mathematics. Students also completed a task­specific questionnaire designed to measure their self­efficacy for successfully completing mathematics problems (five questions).

To measure students’ epistemic and learning beliefs, we adapted items from Schoenfeld’s (1988) and Kloosterman’s (1991) questionnaires. To measure students’ achievement goal orientations, we modified items from Elliot and McGregor’s (2001) achievement goals questionnaire. Previous scales that have been adapted for elementary students have been reliable, valid, and effective in measuring elementary students’ achievement goals in the context of physical education (e.g., Xiang and Lee, 2002). For the questionnaire, students rated each item on a five­point Likert scale ranging from “completely disagree” (a rating of one) to “completely agree” (a rating of five). Each item was anchored with a descriptor to help students interpret the meaning of each number (e.g., two = somewhat disagree, three = don’t know, four = somewhat agree). A sample item from the belief in effort scale is “Students do well in math if they try to figure things out.” A sample item from the belief in certain and simple knowledge scale is “There is only one right answer in math.” A sample item from the mastery goals scale is “It is important for me to understand math.” Finally, a sample item from the perform­ance approach goal scale is “I want to do better than other students in math.” (All items are presented in Appendix 14.A.1.)

To measure self­efficacy, task­specific measures were included to assess whether students felt confident they could correctly solve the problems. Using the guidelines Pajares (1996) specified for measur­ing task­specific self­efficacy, a self­efficacy scale was designed that included five mathematics problems similar to those in the achievement test (sample items are presented in Appendix 14.A.2 for grade four). For each problem, students were given enough time to read each prob­lem but not enough time to actually solve them. After each problem was

Effects of teachers’ beliefs on elementary students448

presented, students rated their self­efficacy using a seven­point Likert scale ranging from “not confident at all” (a rating of one) to “very con­fident” (a rating of seven).

Students’ mathematics achievement. To measure mathematics achieve­ment, two fifteen­item assessment tests were developed, one for each grade. All items were drawn from standardized achievement tests developed specifically for the students in the state from which they were sampled. Six of the items were intended to measure students’ knowledge of place value, addition/subtraction concepts and facts, and multiplication/division concepts and facts. The other nine items were word problems designed to measure students’ knowledge of mathemat­ics vocabulary. Students received one point for each correct answer, for a total of fifteen possible points. (Sample items from the grade four assessment test are presented in Appendix 14.A.3.)

Procedure

Teachers spent fifteen minutes completing a demographics question­naire (e.g., to measure age, number of years’ teaching, etc.) and the DSBQ (Buehl et al., 2002). Once finished, teachers were given writ­ten and verbal directions on how to instruct students to complete the student questionnaires. Approximately one to two weeks later, teach­ers had all participating students complete the student questionnaires. Prior to solving any problems, students filled out the epistemic and learning beliefs and achievement goals questionnaire followed by the self­efficacy questionnaire. Teachers read out instructions to students as well as all items on which they were to rate their agreement or disa­greement. Once students completed the two questionnaires, they were given the mathematics problems to solve on their own. Students spent approximately one hour over the course of two thirty­minute sessions to complete all components.

Results

Since fifty­three students were enrolled in a combined grade four–five class, and to increase sample size, we first explored whether the two groups from grades 4 and 5 could be pooled by examining equivalence of the variances and covariances across groups, as well as factorial invariance. For the first assessment, a multivariate analysis of variance was conducted to examine whether group means across the variables were equivalent. Box’s m test and Levene’s test were also conducted to examine whether the variance­covariance matrix and variances were

Krista R. Muis and Michael J. Foy 449

equivalent across groups. The multivariate results indicated no dif­ference between groups (p = .21). Box’s test of equality of covariances revealed no difference between groups (p = .37), as did Levene’s test for equality of variances (all p > .33).

For the second assessment, we tested for factorial invariance across the two groups. Five steps are involved in testing this property. In the first step, the data are tested separately with each group for fit to the model (Byrne, 1998), in our case, the two dimensions of epistemic and learning beliefs and the two dimensions of achievement goals. If this test is passed, four more tests are applied to establish that no differences exist across groups as a function of the measurement model (see Che­ung and Rensvold, 2002). The fit of the model after adding a constraint is compared to the model at the previous step in terms of differences in Chi­square and comparative fit index (CFI) (Cheung and Rensvold, 2002). A ∆CFI value less than or equal to –.01 indicates that the null hypothesis of invariance should not be rejected (Cheung and Rensvold, 2002). If the fit of the more constrained model is statistically or practi­cally worse than the less constrained model at the prior step, it would be concluded the parameters being constrained differ across groups, that is, they are not invariant. Due to space constraints, we do not detail each of the analyses here, but rather report brief summaries. Specifi­cally, for both grades four and five, fit indices for the measurement models were within a good range (CFIs were .89 and .87, respectively), and changes in fit indices across the remaining four steps were no larger than .01. Accordingly, the two samples were merged.

Data were screened for normality. All teacher and student variables were normally distributed with skewness and kurtosis values within acceptable ranges (e.g., between –1.40 to .33 for skewness, and between –.29 to 1.40 for kurtosis). Means, standard deviations, and reliabil­ity coefficients are presented in Table 14.1. Consistent with previous research, reliability estimates for the subscales were considered reliable with internal consistency α coefficients ranging from .56 to .77.

We used EQS (structural equation modelling software) (Bentler and Wu, 1995) to calculate parameter estimates to test our hypotheses. For graphical simplicity, we show only the factors estimated (ellipses) and omit their respective indicator variables. For all analyses, however, indi­cator variables were used as measures of the latent factors in the model. The hypothesized model displayed in Figure 14.2 is simplified relative to the general model. For achievement, however, since this variable is a direct measure, no underlying variables were used.

We tested the model characterized by Figure 14.2 according to the framework outlined by Baron and Kenny (1986). Beyond testing the

Effects of teachers’ beliefs on elementary students450

relationships between the predictor and mediating variables, two addi­tional structural equation models were necessary. First, predictor vari­ables need to be related to the criterion variables. That is, mediator variables need to be removed from the model. Second, the effect of the predictor variables on the criterion variables need to be signifi­cantly reduced or eliminated when considered jointly with the medi­ator variables. Thus, both paths from the predictor variables to the mediator variables and predictor variables to the criterion variables were included in the model. Based on these criteria, the mediation paths we predicted were supported. Specifically, self­efficacy medi­ated the relationships between a performance approach goal orienta­tion and achievement and a mastery goal orientation and achievement. The final model is presented in Figure 14.2 along with standardized coefficients. Our hypotheses were supported; all paths were statisti­cally detectable.

Based on Byrne’s (1998) recommendations for interpreting fit indi­ces, the estimation of our hypothesized model resulted in a moderate fit, χ 2 (170, N = 131) = 298.18, CFI = .76, and RMSEA = .08. Consist­ent with Biggs’ (1993) and Schommer­Aikins’ (2004) models, as pre­dicted, teachers’ beliefs about the need for effort to learn mathematics were positively related to students’ beliefs about the need for effort to learn mathematics, .10, and positively related to students’ mathematics achievement, .37. Teachers’ beliefs about the integration of information were negatively related to students’ beliefs about the certainty and sim­plicity of mathematics knowledge, –.75, and positively related to stu­dents’ mathematics achievement, .12.

Table 14.1. Means, standard deviations, and alpha coefficients for subscales measured

Mean SD α

Teachers’ belief in effort 5.07 .66 .66Teachers’ belief in integration 6.64 .41 .60Students’ belief in effort 4.54 .55 .71Students’ belief in certain/simple knowledge 3.47 .81 .56Students’ mastery goals 4.55 .69 .77Students’ performance goals 3.92 1.01 .74Students’ self­efficacy 5.16 1.13 .81Students’ achievement 6.55 2.72

Note. Teachers’ beliefs were measured on a seven­point Likert scale. Students’ beliefs and achievement goals were measured on a five­point Likert scale. Students’ self­ efficacy was measured on a seven­point Likert scale. The maximum score possible on students’ mathematics achievement was fifteen.

Krista R. Muis and Michael J. Foy 451

Consistent with achievement goal theory (e.g., Dweck and Leggett, 1988), students’ beliefs about the need for effort to learn mathemat­ics were negatively related to their performance approach goal orienta­tions, –.33, but were positively related to their mastery goal orientations, .67. In contrast, students’ beliefs about the certainty and simplicity of mathematics knowledge were positively related to their performance approach goal orientations, .42, and negatively related to their mas­tery goal orientations, –.59. Finally, as predicted, students’ self­efficacy mediated the negative relationship between a performance approach goal orientation, –.45, and achievement, .39, and mediated the positive relationship between a mastery goal orientation, .19, and achievement. We discuss and elaborate these results in the context of various theor­etical frameworks.

Discussion

We examined whether teachers’ epistemic and teaching and learning beliefs about mathematics influenced students’ epistemic and learning beliefs about mathematics and mathematics achievement, and whether students’ beliefs subsequently influenced their levels of mastery and performance approach goals, self­efficacy, and mathematics achieve­ment. Consistent with Biggs’ (1993) 3P model of teaching and learning and Schommer­Aikins’ (2004) embedded systemic model of epistemic beliefs, teachers’ beliefs were significant predictors of students’ beliefs

Teachers’belief ineffort

Students’belief ineffort

Teachers’belief inintegration

Students’belief incertain/simpleknowledge

Masterygoals

Performancegoals

Self-efficacy Achievement

.10

– .75

– .33

– .59

.67

.42

.19

– .45

.39

.37

.12

Figure 14.2: Final model

Effects of teachers’ beliefs on elementary students452

and student achievement. Moreover, consistent with our hypotheses and achievement goal theory, students’ beliefs subsequently influenced their levels of achievement goals. Finally, self­efficacy mediated rela­tions between achievement goals and achievement. We discuss each of these relations in turn.

Teachers’ and students’ epistemic and learning beliefs

Teachers’ beliefs about the need for effort to learn mathematics posi­tively predicted students’ beliefs about effort and students’ mathemat­ics achievement. Similarly, teachers’ beliefs about the integration of information in mathematics negatively predicted students’ beliefs about the certainty and simplicity of knowledge and positively predicted stu­dents’ achievement. We speculate these results may be a function of the types of statements teachers make when teaching the content, and/or the types of instructional strategies they use to teach. For example, if teachers encourage students to engage in effortful activities when learn­ing mathematics, or use instructional strategies that demonstrate effort is needed to learn, then teacher encouragement and student learning activities may subsequently influence students’ beliefs and, to a greater extent, students’ achievement. Teachers may also relate mathematics content to other domains of learning. For example, understanding the concept of symmetry can be integrated with science content by identi­fying plants by their leaf and flower patterns. The explicit integration of information or connections that teachers make across content areas may subsequently influence students’ beliefs and, to a lesser extent, students’ achievement.

Although we did not observe classroom activities and interac­tions, our hypotheses are consistent with previous research that has demonstrated coherence between teachers’ beliefs and practice (e.g., Brickhouse, 1989; Hammrich, 1998; Hashweh, 1996; Tsai, 2006), and relations between instructional strategies and students’ beliefs (e.g., Hofer, 1999; Schoenfeld, 1988; Stodolsky, 1985). Results from these studies, including ours, also support Muis et al.’s (2006) Theory of integrated domains in epistemology (TIDE) framework. According to Muis et al., interactions with the environment play an important role in the development, modification, and expression of beliefs; that is, indi­viduals actively construct or make meaning of their experiences, and development occurs as a function of one’s interactions with the social world (Baxter Magolda, 2004; Belenky et al., 1986; Bendixen and Rule, 2004; Hofer and Pintrich, 1997). Of particular interest, Muis et al. pro­pose that domain­specific epistemic beliefs are socially constructed and

Krista R. Muis and Michael J. Foy 453

context bound; the context is the instructional environment, which is also embedded in the academic context and the socio­cultural context. Muis et al. hypothesized that the instructional environment shapes stu­dents’ domain­specific epistemic beliefs, which includes grading and school policies and practices (Schoenfeld, 1989).

Similarly, according to Schraw (2001), schools shape and change stu­dents’ epistemic beliefs in a number of ways. Schools may influence beliefs through teacher modeling. That is, consistent with our results, Schraw proposed that teachers’ epistemic beliefs might directly influ­ence students’ epistemic beliefs. Schools may also provide a “training ground” for students to develop critical thinking skills that allow them to think about, use, and modify their own views of knowledge. Accord­ingly, our research lends further support to the notion that the epis­temic climate within which students learn has an important influence on students’ epistemic beliefs; teachers’ beliefs may be one fundamental component for the development of that climate. We discuss the educa­tional implications in the context of teaching mathematics following a brief analysis of the relations between beliefs, achievement goals, self­efficacy, and achievement.

Students’ epistemic and learning beliefs, achievement goals, self-efficacy, and achievement

Hofer and Pintrich (1997) hypothesized that epistemic beliefs may function as implicit theories that may subsequently influence the types of achievement goals students adopt in a given learning situation. Simi­larly, based on a review of theoretical and empirical work, Muis (2007) further elaborated that epistemic beliefs are one component of the cognitive conditions of a task that subsequently influence the types of standards students set for learning, such as time on task, and epistemo­logical standards. According to her model, given that achievement goals are a part of the multifaceted information included in a set of standards, epistemic beliefs also influence achievement goals. Moreover, within the context of achievement goal theory (e.g., Dweck and Leggett, 1988), researchers have also proposed that the types of achievement goals stu­dents set subsequently influence their self­efficacy for engaging in a particular task.

Our results provide support for these predicted relations. We found that the more students believed in the need for effort to learn math­ematics, the higher their mastery goal orientation and the lower their performance approach orientation. Similarly, the more students believed that mathematics knowledge is simple and certain, the higher

Effects of teachers’ beliefs on elementary students454

their performance approach goal orientation and the lower their mas­tery goal orientation. Moreover, students with higher mastery goals also had higher self­efficacy for mathematics problem­solving, whereas stu­dents with higher performance approach goals had lower levels of self­efficacy for mathematics problem­solving.

A positive relationship between mastery goals and self­efficacy and a negative relationship between performance approach goals and self­ efficacy is consistent with the normative goal perspective within the achievement goal framework. Proponents of the normative goal perspec­tive suggest that adoption of mastery goals leads to optimal outcomes, such as higher self­efficacy and achievement, whereas the adoption of performance approach goals leads to less optimal outcomes, such as lower self­efficacy and achievement (Kaplan and Middleton, 2002). In contrast, advocates of the multiple goals perspective propose that adoption of both high mastery and performance approach goal orientations is most beneficial (Harackiewicz et al., 2002; Pintrich, 2000). Tradition­ally, performance approach goals have been theoretically associated with less adaptive patterns of learning and achievement (Ames, 1992). Recently, however, several studies have empirically demonstrated that performance approach goals can be beneficial to learning (e.g., Meece et al., 1988; Wolters et al., 1996) and achievement (e.g., Bouffard et al., 1998; Harackiewicz et al., 2000; Pintrich, 2000). Although results have been mixed across a wide range of studies, our results indicate that in the context of mathematics problem­solving whereby specific beliefs about mathematics are measured, adoption of mastery goals results in higher self­efficacy and higher achievement. In contrast, adoption of performance approach goals results in lower self­efficacy and lower achievement. Much more research is needed, however, to clarify the conditions and contexts within which performance approach goals are beneficial.

Finally, our results are consistent with previous research that has demonstrated a mediating relationship between achievement goals, self­efficacy, and achievement (e.g., VandeWalle et al., 2001). Similar to VandeWalle et al. (2001), we found that self­efficacy mediated the posi­tive relationship between mastery goals and achievement, and medi­ated the negative relationship between performance approach goals and achievement. Given that self­efficacy continues to be one of the larg­est predictors of achievement compared to other cognitive and motiva­tional constructs, such as study skills and achievement goals (Robins et al., 2004), we suggest it is pertinent that educators create classroom cli­mates that foster higher levels of self­efficacy for learning. In the context of our study, two important antecedents to higher levels of self­efficacy

Krista R. Muis and Michael J. Foy 455

were teachers’ epistemic and learning beliefs. We propose that if teach­ers promote effortful learning and integration of information, this may indirectly influence students’ self­efficacy. If students engage in effort­ful learning, which then leads to greater academic success, then this may foster higher levels of self­efficacy (Bandura, 1997). What are the specific educational implications with regard to teaching mathematics and how might they translate into broader educational implications? We discuss these next.

Specific educational implications for teaching mathematics

As previously discussed, students at all levels of education espouse non­constructivist beliefs about mathematics (see Muis, 2004), and several calls have been made for instructional reform to address this issue. As Muis (2004) identified in her review, all too often teachers rely on “drill and kill” techniques to teach students mathematics. As the NCTM (2006) highlights, this is not a particularly useful strategy to help foster students’ understanding of mathematics, and likely has negative effects on their beliefs about mathematics as well as their per­formance (Muis, 2004). It is of our opinion that mathematics stands at a disadvantage to other content areas due to the instructional strate­gies and assessment techniques that teachers use to teach mathematics and to measure student learning, which has serious implications for the development of students’ epistemic beliefs. Why might mathematics be at a disadvantage?

Several previous observational studies (e.g., Schoenfeld, 1988; Stodolsky et al., 1991) have revealed that, when teaching mathematics, teachers rely on more traditional approaches to instruction. Traditional approaches include teaching that focuses on speed, accuracy, and mem­orization of rules and procedures presented by the teacher and prac­ticed in isolation. As Muis (2004) noted in her review, these types of strategies are associated with beliefs that learning is quick, there is only one right answer, success requires innate ability, mathematical knowl­edge is unchanging and consists of isolated pieces of information, and the teacher is the source by which to justify mathematical knowledge. We argue that one potential reason for reliance on these instructional approaches is the nature of mathematics knowledge (see also Muis et al., 2006). Specifically, the underlying structure of mathematics knowledge is assumed to be formal, and consists of syntactical rules and elements (Triadafillidis, 1998).

According to Triadafillidis (1998), Cartesianism, Platonism, and Formalism continue as the dominant epistemologies in mathematics

Effects of teachers’ beliefs on elementary students456

education. To Plato, mathematics knowledge was a gateway to phi­losophy and an attempt to discover eternal and universal principles. The teleological dimension of doing mathematics was realized in knowing mathematics and the acquisition of scientific truth. Simul­taneously, mathematical theories were elaborated around a set of axi­oms. Thus began mathematical Formalism, the practice of following syntactico­grammatical canons and protocols that dictated the proper use of symbols. Platonism and Formalism manifest themselves in the instructional techniques teachers use to teach mathematics. For exam­ple, researchers in mathematics education have shown that US class­rooms are typically concerned with student mastery of algorithmic procedures. The norm in mathematics instruction is teacher explana­tion of algorithmic procedures followed by student practice through individual seatwork exercises, exercises that have well­defined proce­dures and answers. Cooperative learning is uncommon, and exposure to content is presented in a fairly well­defined sequence (Stodolsky et al., 1991).

We argue that the structure of mathematics itself fosters a control­led and sequenced step­by­step approach to teaching mathematics; teachers rarely engage students in multiple ways to solve problems. Moreover, we suggest that the assessment tools teachers use also foster non­constructivist epistemic beliefs. That is, we question how teach­ers are to communicate the complexity of mathematics knowledge and foster student understanding of underlying constructs when the assessment techniques are primarily paper­and­pencil based (NCTM, 2006). What alternative assessments are being used in the mathematics classrooms?

When the focus of assessments is on finding the correct answer, which is achieved via one approach to solving a problem, it is unlikely that students’ epistemological development will progress toward more con­structivist views. Given the relationship between teachers’ beliefs and instructional approaches (e.g., Hashweh, 1996; Tsai, 2006), and results from our study, we argue that it is pertinent not only to encourage teachers and preservice teachers to use more constructivist techniques, but to also assess their beliefs about mathematics knowledge and foster their epistemological development. For example, teachers could assess students’ beliefs by engaging students in discourse about “What is mathematics?” and “What does it mean to do mathematics?” Fostering epistemological development in teachers may also help to foster episte­mological development in students. How might this be achieved? We address this next in our broader educational implications.

Krista R. Muis and Michael J. Foy 457

Broader educational implications

Previous studies have found coherence between teachers’ beliefs and instructional practices (e.g., Tsai, 2006), and between instructional practices and students’ epistemic and learning beliefs (e.g., Hammer and Elby, 2002). Our research adds to this literature by examining directly the relationship between teachers’ and students’ epistemic and learning beliefs. We discuss three important broader educational implications of our research. First, in the context of preservice teacher training, greater attempts must be made to develop training curricula that provide a comprehensive analysis of the epistemic foundations of teaching and that highlight the central role of teachers’ beliefs in student learning. Teachers should be made explicitly aware of their beliefs and how those beliefs can influence instructional approaches, students’ motivation, and student achievement. Similarly, attempts must be made to train preservice teachers in more constructivist approaches to learning that parallel constructivist beliefs about knowledge and learning. Unfortu­nately, current research suggests that, while constructivist conceptions of teaching and learning have been implemented into teacher training programs, preservice teachers continue to espouse more traditionalist, teacher­centered beliefs (e.g., Hartzell and Muis, 2006; Schraw and Olafson, 2002). As previous research has demonstrated, these types of teacher­centered beliefs result in specific types of instructional strat­egies that negatively influence students’ motivation and performance (e.g., Daniels and Shumow, 2003). Accordingly, not only do teacher­training programs need to include these components into the training curricula, more in­service support is necessary to ensure correct and continued implementation in the classrooms.

Second, although some teachers espouse more constructivist beliefs, research has also demonstrated that teachers’ instructional practices are not always consistent with their beliefs (e.g., Levitt, 2001; Schraw and Olafson, 2002; White, 2000). For example, Schraw and Olafson (2002) found that teachers had to use state­mandated curriculum and instruc­tional strategies. When curriculum is scripted, teachers may be less likely to implement strategies that result in better learning outcomes. In addi­tion, given the No Child Left Behind Act (NCLB) in the United States, a major focus in schools today is performance on standardized achieve­ment tests. State­mandated curriculum coupled with high­stakes testing may lead educators to focus solely on student performance, rather than on mastery of the content. In this case, regardless of what teachers’ beliefs may be, the classroom climate may become performance oriented, which may subsequently result in negative outcomes (e.g., Urdan, 2004).

Effects of teachers’ beliefs on elementary students458

Finally, as Gunzenhauser (2003) argues, high­stakes testing may lead to a “default” philosophy of education that holds a narrow bundle of knowledge and skills in high regard. He defines a default philosophy as a vision of education that results from a lack of reflective and engaged dialogue among educators and school communities about their goals and practices. With a high­stakes testing focus, the default philosophy predominates which includes an inordinate focus on the tests them­selves. This creates a classroom climate wherein the emphasis is on test preparation, rather than on the content itself (Patterson, 2002). As several studies have demonstrated, conversations about the mean­ing and value of education do not take place without performance on standardized tests taking central stage (e.g., Linn, 2000; McGehee and Griffith, 2001; Murillo and Flores, 2002). Gunzenhauser notes this type of environment is not likely to foster meaningful dialogue, and it may become more of a challenge to incorporate other philosophies of education, such as constructivist approaches. Restricting teachers’ opportunities to implement alternative approaches may limit possibili­ties for educational reform.

In line with our recommendations, over the past two decades the National Science Council (NSC, 2006) and the National Council of Teachers of Mathematics (NCTM, 2006) have made calls for radi­cal shifts in mathematics instruction. Because of the growing concern among mathematics educators regarding students’ beliefs and how they impact learning, the Standards (NCTM, 2006) suggest that the assess­ment of teachers’ and students’ beliefs about mathematics is a crucial component of the general assessment of knowledge of mathematics. One recommendation is for teachers to create more constructivist­oriented classroom environments. They suggest that teachers be trained to develop classroom environments that focus on providing interesting and relevant problems. The NSC and NCTM also suggest that teachers shift the focus of the activities away from them (e.g., a teacher­centered approach to teaching) to the students by providing students opportuni­ties to explore, work in small groups, and to emphasize understand­ing of concepts rather than computation. However, with pressures from NCLB and in a time when teacher and school accountability is the primary focus, implementation of constructivist approaches may be thrown to the wayside. Accordingly, we question the value of such high­stakes testing and suggest alternative methods of accountability be developed and implemented. What these may be should be the sub­ject of future discussion.

Krista R. Muis and Michael J. Foy 459

Limitations and future directions

Our results demonstrate that teachers’ epistemic and learning beliefs about mathematics influence students’ beliefs and achievement. Why are these related? We interpreted our results in the context of Biggs’ (1993) 3P model of teaching and learning and Schommer­Aikins’ (2004) embedded systemic model of epistemic beliefs. Although our hypotheses were supported, we can only provide speculative discussion as to why these constructs are related. What is needed is research that includes direct observations of the types of strategies and assessment techniques teachers implement and employ in their classrooms, as well as interviews with teachers and students to gain a more in­depth under­standing of their epistemic and teaching and learning beliefs. Teach­ers’ instructional and assessment strategies may serve as an important mediating variable between teachers’ and students’ beliefs. Further­more, research is needed to examine whether this relationship holds in contexts where constraints prevent teachers from enacting their beliefs. We can only hope that in such instances when teachers espouse more constructivist beliefs, their beliefs continue to play an important role in the development of students’ beliefs.

R EF ER ENCES

Ames, C. (1992). Classrooms: Goals, structures, and student motivation. Jour-nal of Educational Psychology, 84, 261–71.

Ames, C. and Archer, J. (1988). Achievement goals in the classroom: Stu­dents’ learning strategies and motivation processes. Journal of Educational Psychology, 80, 260–70.

Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W. H. Freeman and Company.

Baron, R. M. and Kenny, D. A. (1986). The moderator­mediator variable distinc­tion in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–82.

Baxter Magolda, M. B. (2004). A constructivist conceptualization of epistemo­logical reflection. Educational Psychologist, 39(1), 31–42.

Belenky, M., Clinchy, B., Goldberger, N., and Tarule, J. (1986). Women’s ways of knowing: The development of self, voice, and mind. New York: Basic Books.

Bendixen, L. D. and Rule, D. C. (2004). An integrative approach to personal epistemology: A guiding model. Educational Psychologist, 39(1), 69–80.

Benson, G. D. (1989). Epistemology and science curriculum. Journal of Curriculum Studies, 21, 329–44.

Bentler, P. M. and Wu, E. J. C. (1995). EQS for windows: User’s guide. Encino, CA: Multivariate Software, Inc.

Effects of teachers’ beliefs on elementary students460

Biggs, J. B. (1993). From theory to practice: A cognitive systems approach. Higher Education Research and Development, 12, 73–85.

Bouffard, T., Vezeau, C., and Bordeleau, L. (1998). A developmental study of the relation between combined learning and performance goals and students’ self­regulated learning. Journal of Educational Psychology, 68, 309–19.

Bråten, I. and Olaussen, B. S. (2005). Profiling individual differences in student motivation: A longitudinal cluster­analytic study in different aca­demic contexts. Contemporary Educational Psychology, 30, 359–96.

Brickhouse, N. W. (1989). The teaching of the philosophy of science in second­ary classrooms: Case studies of teachers’ personal theories. International Journal of Science Education, 11, 437–49.

Buehl, M. M. (2003). At the crossroads: Exploring the intersection of epistemological beliefs, motivation, and culture. Paper presented at the annual meeting of the American Educational Research Association, Chicago (April).

Buehl, M. M., Alexander, P. A., and Murphy, P. K. (2002). Beliefs about schooled knowledge: Domain specific or domain general? Contemporary Educational Psychology, 27, 415–49.

Burr, J. E. and Hofer, B. K. (2002). Personal epistemology and theory of mind: Deciphering young children’s beliefs about knowledge and know­ing. New Ideas in Psychology, 20, 199–224.

Button, S., Matthieu, J., and Zajac, D. (1996). Goal orientation in organiza­tional behavior research: A conceptual and empirical foundation. Organi-zational Behavior and Human Decision Processes, 67, 26–48.

Byrne, B. M. (1998). Structural equation modeling with LISREL, PRELIS, and SIMPLIS. Mahwah, NJ: Lawrence Erlbaum Associates.

Carter, C. S. and Yackel, E. (1989). A constructivist perspective on the relation-ship between mathematical beliefs and emotional acts. Paper presented at The Annual Meeting of the American Educational Research Association, San Francisco, CA.

Cheung, G. W. and Rensvold, R. B. (2002). Evaluating goodness­of­fit indexes for testing measurement invariance. Structural Equation Modeling, 9, 233–55.

Daniels, D. H. and Shumow, L. (2003). Child development and classroom teaching: A review of the literature and implications for educating teach­ers. Applied Developmental Psychology, 23, 495–526.

Doyle, W. (1988). Work in mathematics classes: The context of students’ think­ing during instruction. Educational Psychologist, 23, 167–80.

Duschl, R. A. and Wright, E. (1989). A case study of high school teachers’ deci­sion making models for planning and teaching science. Journal of Research in Science Teaching, 26, 547–59.

Dweck, C. S. and Leggett, E. L. (1988). A social­cognitive approach to motiva­tion and personality. Psychological Review, 95, 256–73.

Elliot, A. J. and McGregor, H. (2001). A 2 x 2 achievement goal framework. Journal of Personality and Social Psychology, 80, 501–19.

Frank, M. L. (1988). Problem solving and mathematical beliefs. Arithmetic Teacher, 35, 32–4.

Krista R. Muis and Michael J. Foy 461

Franke, M. L. and Carey, D. A. (1997). Young children’s perceptions of math­ematics in problem­solving environments. Journal for Research in Math-ematics Education, 28, 8–25.

Garofalo, J. (1989). Beliefs and their influence on mathematical performance. Mathematics Teacher, 82, 502–5.

Gunzenhauser, M. G. (2003). High stakes testing and the default philosophy of education. Theory into Practice, 42, 51–8.

Hammer, D. H. and Elby, A. (2002). On the form of personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (169–90). Mahwah, NJ: Lawrence Erlbaum Associates.

Hammrich, P. L. (1998). Cooperative controversy challenges elementary teacher candidates’ conceptions of the “nature of science.” Journal of Ele-mentary Science Education, 10, 50–65.

Harackiewicz, J. M., Barron, K. E., Pintrich, P. R., Elliot, A. J., and Thrash, T. M. (2002). Revision of achievement goal theory: Necessary and illumi­nating. Journal of Educational Psychology, 94(3), 638–45.

Harackiewicz, J. M., Barron, K. E., Tauer, J. M., Carter, S. M., and Elliot, A. J. (2000). Short­term and long­term consequences of achievement goals: Predicting interest and performance over time. Journal of Educa-tional Psychology, 92, 316–30.

Hartzell, S. A. and Muis, K. R. (2006). Pre-service teachers’ belief adherence to constructivist theory. Paper presented at the annual meeting of the Ameri­can Educational Research Association, San Francisco, CA.

Hashweh, M. Z. (1996). Effects of science teachers’ epistemological beliefs in teaching. Journal of Research in Science Teaching, 33, 47–63.

Higgins, K. M. (1997). The effect of year­long instruction in mathematical problem solving on middle­school students’ attitudes, beliefs, and abili­ties. Journal of Experimental Education, 66, 5–28.

Hofer, B. K. (1999). Instructional context in the college mathematics class­room: Epistemological beliefs and student motivation. Journal of Staff, Program, and Organizational Development, 16, 73–82.

Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learn­ing. Review of Educational Research, 67, 88–140.

(2002). Personal epistemology: The psychology of beliefs about knowledge and knowing. Mahwah, NJ: Lawrence Erlbaum Associates.

Johnston, P., Woodside­Jiron, H., and Day, J. (2001). Teaching and learning literate epistemologies. Journal of Educational Psychology, 93(1), 223–33.

Kaplan, A. and Middleton, M. J. (2002). Should childhood be a journey or a race? A response to Harackiewicz et al. (2002). Journal of Educational Psychology, 94, 646–8.

Kloosterman, P. (1991). Beliefs and achievement in seventh­grade mathemat­ics. Focus on Learning Problems in Mathematics, 13, 3–15.

Lambert, N. M. and McCombs, B. L. (1998). How students learn: Reforming schools through learner-centered education. Washington, DC: American Psy­chological Association.

Effects of teachers’ beliefs on elementary students462

Lampert, M. (1990). When the problem is not the question and the solution is not the answer: Mathematical knowing and teaching. American Educa-tional Research Journal, 27, 29–63.

Lederman, N. G. (1992). Students’ and teachers’ conceptions of the nature of science: A review of the research. Journal of Research in Science Teaching, 29, 331–59.

(1999). Teachers’ understanding of the nature of science and classroom prac­tice: Factors that facilitate or impede the relationship. Journal of Research in Science Teaching, 36, 916–29.

Lederman, N. G. and Zeidler, D. L. (1987). Science teachers’ conceptions of the nature of science: Do they really influence teaching behavior? Science Education, 71, 721–34.

Levitt, K. E. (2001). An analysis of elementary teachers’ beliefs regarding the teaching and learning of science. Science Education, 86, 1–22.

Linder, C. J. (1992). Is teacher­reflected epistemology a source of concep­tual difficulty in physics? International Journal of Science Education, 12, 111–21.

Linn, R. L. (2000). Assessments and accountability. Educational Researcher, 29, 4–16.

McGehee, J. J. and Griffith, L. K. (2001). Large scale assessments combined with curriculum alignment: Agents of change. Theory into Practice, 40, 137–44.

Meece, J. L., Blumenfeld, P., and Hoyle, R. (1988). Students’ goal orientations and cognitive engagement in classroom activities. Journal of Educational Psychology, 80, 514–23.

Muis, K. R. (2004). Personal epistemology and mathematics: A critical review and synthesis of research. Review of Educational Research, 74, 317–77.

(2007). The role of epistemic beliefs in self­regulated learning. Educational Psychologist, 42(3), 173–90.

Muis, K. R., Bendixen, L. D., and Haerle, F. C. (2006). Domain­generality and domain­specificity in personal epistemology research: Philosophical and empirical reflections in the development of a theoretical framework. Educational Psychology Review, 18(1), 3–54.

Murillo, E. G. and Flores, S. Y. (2002). Reform by shame: managing the stigma of labels in high stakes testing. Educational Foundations, 16, 93–108.

National Council of Teachers of Mathematics (NCTM) (2006). Principles and standards for school mathematics. Reston, VA: NCTM.

National Science Board (2006). National Science Board commission on 21st cen-tury education in science, technology, engineering, and mathematics. Reston, VA: NSC.

Neber, H. and Schommer­Aikins, M. (2002). Self­regulated learning with highly gifted students: The role of cognitive, motivational, epistemologi­cal, and environmental variables. High Ability Studies, 13, 59–74.

Pajares, F. (1992). Teachers’ beliefs and educational research: Cleaning up a messy construct. Review of Educational Research, 62, 307–32.

(1996). Self­efficacy beliefs in academic settings. Review of Educational Research, 66, 543–78.

Krista R. Muis and Michael J. Foy 463

Patterson, J. A. (2002). Exploring reform as symbolism and expression of belief. Educational Foundations, 16, 55–75.

Peterman, F. (1993). Staff development and the process of changing: A teach­er’s emerging beliefs about learning and teaching. In K. Tobin (Ed.), The practice of constructivism in science education (227–45). Hillsdale, NJ: Law­rence Erlbaum Associates.

Pintrich, P. R. (1990). Implications of psychological research on student learning and college teaching for teacher education. In W. R. Houston (Ed.), Hand-book of research on teacher education (826–57). New York: Macmillan.

(2000). Multiple goals, multiple pathways: The role of goal orientation in learning and achievement. Journal of Educational Psychology, 92, 544–55.

(2002). Future challenges and directions for theory and research on per­sonal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing (389–414). Mahwah, NJ: Lawrence Erlbaum Associates.

Robins, S. B., Lauver, K., Le, H., Davis, D., and Langley, R. (2004). Do psychosocial and study skills factors predict college outcomes? A meta­ analysis. Psychological Bulletin, 130, 261–88.

Schoenfeld, A. H. (1988). When good teaching leads to bad results: The dis­asters of “well­taught” mathematics classes. Educational Psychologist, 23, 145–66.

(1989). Exploration of students’ mathematical beliefs and behavior. Journal for Research in Mathematics Education, 20, 338–55.

Schommer, M. (1990). Effects of beliefs about the nature of knowledge on comprehension. Journal of Educational Psychology, 82(3), 498–11.

(1998). The role of adults’ beliefs about knowledge in school, work, and everyday life. In M. C. Smith and T. Pourchot (Eds.), Adult learning and development: Perspectives from educational psychology (127–43). Mahwah, NJ: Lawrence Erlbaum Associates.

Schommer­Aikins, M. (2004). Explaining the epistemological belief sys­tem: Introducing the embedded systemic model and coordinated research approach. Educational Psychologist, 39, 19–29.

Schommer­Aikins, M., Duell, O. K., and Hutter, R. (2005). Epistemologi­cal beliefs, mathematical problem­solving, and academic performance of middle school students. The Elementary School Journal, 105, 289–304.

Schraw, G. (2001). Current themes and future directions in epistemological research: A commentary. Educational Psychology Review, 13, 451–65.

Schraw, G. and Olafson, L. (2002). Teachers’ epistemological world views and educational practices. Issues in Education, 8, 99–148.

Stodolsky, S. S. (1985). Telling math: Origins of math aversion and anxiety. Educational Psychologist, 20, 125–33.

Stodolsky, S. S., Salk, S., and Glaessner, B. (1991). Student views about learn­ing math and social studies. American Educational Research Journal, 28, 89–116.

Tobin, K. (1993). Constructivist perspectives on teacher learning. In K. Tobin (Ed.), The practice of constructivism in science education (215–26). Hillsdale, NJ: Lawrence Erlbaum Associates.

Effects of teachers’ beliefs on elementary students464

Triadafillidis, T. A. (1998). Dominant epistemologies in mathematics educa­tion. For the Learning of Mathematics, 18, 21–7.

Tsai, C.­C. (2002). Nested epistemologies: Science teachers’ beliefs of teach­ing, learning and science. International Journal of Science Education, 24, 771–83.

(2006). Teachers’ scientific epistemological views: The coherence with instruction and students’ views. Science Education, 91, 222–43.

Urdan, T. (2004). Using multiple methods to assess students’ perceptions of classroom goal structures. European Psychologist, 9, 222–31.

VandeWalle, D., Cron, W. L., and Slocum, J. W. Jr. (2001). The role of goal orientation following performance feedback. Journal of Applied Psychology, 86, 629–40.

White, B. C. (2000). Pre­service teachers’ epistemology viewed through per­spectives on problematic classroom situations. Journal of Education for Teaching, 26(3), 279–306.

Wilcox­Herzog, A. (2002). Is there a link between teachers’ beliefs and behav­iors? Early Education and Development, 13(1), 81–106.

Wolters, C., Yu, S., and Pintrich, P. R. (1996). The relation between goal ori­entation and students’ motivational beliefs and self­regulated learning. Learning and Individual Differences, 8, 211–38.

Xiang, P. and Lee, A. (2002). Achievement goals, perceived motivational cli­mate, and students self­reported mastery behaviors. Research Quarterly for Exercise and Sport, 73, 58–65.

A PPEN DI X 14.A .1

For each sentence, circle the number that best describes how you feel about math. There are no right or wrong answers. Use these numbers to help you.

1 = completely disagree2 = disagree3 = don’t know4 = agree5 = completely agree

1. There is more than one way to do math problems. 1 2 3 4 5

2. I see how math is connected to other subjects. 1 2 3 4 5

3. When I learn something new in math, I try to connect it to what I already know.

1 2 3 4 5

4. Some people just can’t do math. 1 2 3 4 5

5. Math is related to things outside of school. 1 2 3 4 5

6. There is only one right answer in math. 1 2 3 4 5

Krista R. Muis and Michael J. Foy 465

7. Trying hard is really important when it comes to learning math.

1 2 3 4 5

8. I like hard problems in math, even when I don’t know how to solve them.

1 2 3 4 5

9. I am happy when I am the only one who got an answer right in math.

1 2 3 4 5

10. I value what I learn in math. 1 2 3 4 5

11. Students do well in math if they try to figure things out. 1 2 3 4 5

12. I want to learn as much as I can in math. 1 2 3 4 5

13. I want to do better than other students in math. 1 2 3 4 5

14. I want to get the highest grade in math. 1 2 3 4 5

15. It is important for me to understand math. 1 2 3 4 5

A PPEN DI X 14.A .2

Problem 1:

Calculate the following question.

23 + 14 – 5 =__________

Problem 2:

Calculate and fill in the answer.

If x + 3 = 14, then x =__________

Problem 3:

Look at the polygon below.

14 in.

5 in.

5 in.

7 in. 7 in.

20 in.

What is the perimeter of the polygon?

Effects of teachers’ beliefs on elementary students466

Problem 4:Megan wants to share some cookies with her friends. She counts out the cookies into groups of 6. Which list shows numbers in groups of 6?A. 7, 14, 21, 28B. 5, 15, 25, 30C. 6, 12, 18, 24D. 14, 23, 28, 35

Problem 5:There are 48 bicycles in the school’s bicycle rack. There are 18 BMX bikes, and the rest are mountain bikes. How many mountain bikes are there?(Note: Each problem was presented on a separate page.)Once students read each problem, they rated their self­efficacy on the following scale for each problem:

1. How confident are you that you could correctly do Problem 1?

1 2 3 4 5 6 7Not confident at all

Not sure Very confident

A PPEN DI X 14.A .3MATHEMATICS ASSESSMENT GRADE 4

1. What is the sum of 2 and 4?__________

Calculate the following questions.2. 13 + 68 – 3 =__________3. 8 × 7 =__________4. 40 ÷ _________ = 8

Fill in the blank with less than (<), greater than (>) or equal to (=) to make the comparison true.5. 0.8 __________ 0.09

Krista R. Muis and Michael J. Foy 467

6. Fill in the columns for each figure pictured in the chart below.

How many pairs ofopposite sides areparallel?

How many pairs ofopposite sides arecongruent?

How many anglesare right angles?

Calculate and fill in the answer.7. If x + 3 = 8, then x =__________

Effects of teachers’ beliefs on elementary students468

Pet Survey

30 percent

50 percent

10 percent

10 percent

Cats DogsNoneOther

How many students own a dog?__________Solve the following problems. Show all of your work.

9. There are 36 bicycles in the school’s bicycle rack. There are 24 BMX bikes, and the rest are mountain bikes. How many mountain bikes are there?

10. Maria has $48.00. A CD costs $6.00. How many CDs can she buy?

11. Amy wants to share some marbles with her friends. She counts out the marbles into groups of 7. Which list shows numbers in groups of 7?A. 7, 14, 21, 28B. 7, 15, 21, 28C. 14, 22, 28, 35D. 14, 23, 28, 35

12. Sam is ordering hot dogs for the school picnic. The table below shows the number of students in each grade.

Study the following graph, then fill in the answer.8. The circle graph below shows how 100 students answered a survey about what kind of pet they owned.

Krista R. Muis and Michael J. Foy 469

Grade Number of students

First 270Second 320Third 290

If each student has 1 hot dog, which is a reasonable number of hot dogs for Sam to order?A. 700B. 800C. 900D. 1000

Look at the polygon below.

3 in.

3 in.

5 in. 5 in.

16 in.

10 in.

13. What is the perimeter of the polygon?A. 41 in.B. 43 in.C. 44 in.D. 48 in.

14. Clear Lake Elementary students are collecting cans to recycle. They collected 315 cans on Monday, 278 cans on Tuesday, and 352 cans on Wednesday. About how many total cans did the students collect on these three days?

15. Which of the following situations requires an exact answer?A. putting some popcorn in a bowlB. guessing the number of raisins in a boxC. estimating the number of kids on a playgroundD. counting the number of students on a bus when coming home from

a field trip

470

15 Teachers’ articulation of beliefs about teaching knowledge: conceptualizing a belief framework

Helenrose Fives Montclair State University

Michelle M. Buehl George Mason University

Teaching is doing whatever it takes to help students learn concepts, patterns, and systems. It is to be knowledgeable about: methods, instruction, curriculum. And using that knowledge to guide students to become inquisitive, active members of learning communities.

(ID 421, practicing teacher)

The remarks above describe one teacher’s perspective on teaching. An examination of this quote garners insights and inferences regarding the practices this teacher may use in her classroom. The perspective that teaching requires doing “whatever it takes,” knowledge about a variety of topics, and the ability to use knowledge in multiple ways and settings, suggests that teaching is a very complex task. Thus, the beliefs that underscore this perspective on teaching must be equally complex. In this chapter we explore how preservice and practicing teachers articulate their beliefs about teaching knowledge. By teach­ing knowledge we mean any knowledge used to facilitate the practice of teaching.

Researchers and teacher educators have explored and examined teachers’ beliefs for decades (Pajares, 1992; Kagan, 1992). Teachers’ beliefs have been examined relative to a number of topics or constructs including motivation (e.g., Stipek et al., 2001), adolescent development (e.g., Buchanan et al., 1990), developmental skills and play (e.g., Kem­ple, 1996; Kowalski et al., 2001), constructivism (e.g., Holt­Reynolds, 2000; McLachlan­Smith and St­George, 2000), instructional practices (e.g., Borko et al., 2000; Lawless and Smith, 1997; Sahin et al., 2002), classroom management and control beliefs (e.g., Weinstein, 1998), diversity (e.g., McAllister and Irvine, 2002; Pohan, 1996), bilingual education (Flores, 2001; Johnson, 2000), special education (e.g., Jordan and Stanovich, 2003; Mantzicopoulos and Neuharth­Prichett, 1998) and urban education (Roderick, 1994). Additionally, there have been

Helenrose Fives and Michelle M. Buehl 471

forays into examining teachers’ beliefs about knowledge and know­ing (i.e., their epistemic beliefs; e.g., Brownlee, 2004; Flowerday and Schraw, 2000).

Similarly, several theorists, including Shulman (1987), Elbaz (1983), and Grossman (1990) have attempted to articulate the knowledge base for teaching. That is, what is it that teachers know that allows them to be successful in facilitating student learning? Although these two strands of research, teachers’ beliefs and teachers’ knowledge, have added greatly to our understanding of teaching and learning, the lack of over­lap between these fields is surprising. For instance, studies of preserv­ice teacher beliefs about knowledge have tended to address their general knowledge beliefs (e.g., Ravindran et al., 2005), focused on their beliefs about knowledge in specific academic domain (e.g., mathematics: Gill et al., 2004), or examined the domain generality­specificity of teachers’ knowledge (e.g., Olafson and Schraw, 2006). Discussions of teacher knowledge rarely query teachers about the knowledge that they need for teaching and few probe teachers’ beliefs about the nature of such knowledge.

To address this issue, we have undertaken an examination of teach­ers’ beliefs that spans these fields of study and connects to the existing literature on students’ knowledge beliefs (Buehl and Fives, 2009; Fives and Buehl, 2005; Fives and Buehl, 2008). In this work, we have devel­oped a framework for understanding teachers’ beliefs that is grounded in previous research yet reflective of the perspectives held by preservice and practicing teachers. Through our examination of their responses and attempts to develop a framework that integrates the voices of teach­ers with the research literature, we have come to realize the complexity of these beliefs. Herein, we overview our framework, which is the cul­mination of several studies, and highlight critical methodological issues that have arisen for us in our attempts to understand these beliefs. We also acknowledge the current theoretical and empirical literature that has influenced our thinking and we relate our emergent framework back to this literature.

Developing a framework of teachers’ beliefs about teaching knowledge

Theoretical influences

We began this line of research in order to explore teachers’ beliefs about teaching knowledge in relation to various issues and constructs related to teacher education, development, retention (e.g., teachers’ sense of

Teachers’ articulation of beliefs about teaching472

efficacy, burnout, mentoring), and practice. Various bodies of litera­ture have informed our thinking including research related to personal epistemology, teacher knowledge and beliefs, motivation, and implicit beliefs. Because extensive and well­written reviews are available else­where (e.g., Calderhead, 1996; Dweck and Molden, 2005; Eccles and Wigfield, 2002; Hofer and Pintrich, 1997; Oser and Baeriswyl, 2001; Pajares, 1992; Richardson, 1996; Wigfield et al., 2006; Woolfolk­Hoy et al., 2006), we do not provide a detailed discussion of these works. Instead, we offer a brief overview of these literatures as they relate to our current work.

Teacher beliefs

Drawing from the literature on teachers’ beliefs, we hold that individuals possess beliefs (i.e., “psychologically held understandings, premises, and propositions about the world that are felt to be true,” Richardson, 1996, p. 103) that guide and influence their actions. Individuals may be explicitly aware of their beliefs such that they consider and examine them regularly. However, it is likely that many beliefs that individuals hold are implicit and, thus, rarely articulated or examined (e.g., Dweck et al., 1995). Examples of possible implicit beliefs include beliefs about ability, morality, knowledge, and the self. Drawing from Dweck’s work on implicit theories about intelligence as well as Schommer’s (1990) belief dimension related to students’ beliefs about one’s ability to learn how to learn, we were interested in teachers’ beliefs about the ability to learn how to teach.

Personal epistemology

Additionally, we, and others as evidenced by this edited volume, are particularly intrigued by beliefs about the nature of knowledge and knowing, specifically how knowledge is defined, constructed, justified, and stored (Hofer, 2002). Thus, our work is informed by various aspects of theory and research related to personal epistemology. Evidence sug­gests that individuals may adopt different perspectives with respect to how one evaluates knowledge or comes to know something and that these perspectives exist along a developmental continuum (e.g., lev­els of reflective thinking: King and Kitchener, 1994, 2002; relativist, multiplist, and evaluativist stances: Kuhn, 1991; Kuhn and Weinstock, 2002). Other research indicates that individuals may possess beliefs about different dimensions of knowledge such as the certainty, simplic­ity, or source of knowledge (i.e., a multidimensional view of knowledge

Helenrose Fives and Michelle M. Buehl 473

beliefs; e.g., Buehl and Alexander, 2005; Hofer, 2000; Schommer, 1990).

Much of the current work in this area has been conducted with stu­dents. However, some research has explicitly focused on practicing or preservice teachers’ beliefs about knowledge (e.g., Gill et al., 2004; Ravindran et al., 2005; Schraw and Olafson, 2006) and references to teachers’ beliefs about knowledge are often implicit in discussions of beliefs about teaching and learning (e.g., Holt­Reynolds, 2000). For instance, Ravindran et al. (2005) found that preservice teachers’ beliefs about the simplicity of knowledge was related to shallow levels of cog­nitive processing. Sinatra and Kardash (2004) found that preservice teachers’ beliefs about knowledge (i.e., beliefs about the complexity of knowledge and the speed of knowledge acquisition) predicted their openness to a new metaphor for teaching (e.g., teaching as persua­sion). In another investigation, Brownlee et al. (2001) examined how preservice teachers’ knowledge beliefs changed over the course of a year­long teaching program. The program was explicitly designed to assist preservice teachers’ reflection on their beliefs about knowledge. More recently, Brownlee and Berthelsen (2006) offered a “conceptual framework that related personal epistemological beliefs and learn­ing outcomes in early childhood teacher education programs” (p. 18). Additionally, Patrick and Pintrich (2001) discussed the role of teach­ers’ epistemic beliefs and motivation in the conceptual change process and highlighted the need for beliefs to be challenged, confronted, and openly discussed. Such endeavors are critical to understanding teach­ers’ beliefs about knowledge and improving teachers’ practices, motiva­tion, and development. However, we feel that greater specificity and a more fine­grained analysis may be needed to clearly understand these beliefs and how they might be changed in applied settings.

Individuals’ beliefs about knowledge and knowing are believed to vary depending on the domain or body of knowledge under consideration (e.g., Buehl and Alexander, 2001, 2006; Hofer, 2000, 2006; Muis et al., 2006; Op ’t Eynde et al., 2006). In previous investigations, researchers have conducted in­depth explorations of students’ beliefs about domains such as mathematics (e.g., de Corte et al., 2002; Schoenfeld, 1992) and science (e.g., diSessa, 1993; Hammer, 1994). Additionally, other works have compared individuals’ beliefs about knowledge in different aca­demic domains (e.g., history and mathematics: Buehl et al., 2002; psy­chology and science: Hofer, 2000). Teachers’ beliefs about knowledge, teaching, and learning relative to different academic domains have also been explored. For instance, Gill et al. (2004) found preservice teach­ers’ implicit and explicit epistemic beliefs about mathematics changed

Teachers’ articulation of beliefs about teaching474

as a result of a text­based intervention. In his examinations of teachers’ epistemological views of science, Tsai (2002, 2007) found that teacher’s views of science are related to their teaching beliefs and practices.

Investigations such as Gill et al.’s (2004) and Tsai’s (2007), appear to be motivated, in part, by a desire to understand teachers’ learning and teaching practices relative to a specific academic domain (e.g., mathematics, science). In contrast, Olafson and Schraw (2006) have engaged in a line of research that explores the potential domain gener­ality of teachers’ knowledge beliefs. These researchers concluded that teachers hold multiple beliefs and engage in teaching practices that are both domain­general and domain­specific. We approach the study of teachers’ beliefs from a perspective that recognizes the multiple beliefs described by Olafson and Schraw (2006), that is, that teachers’ beliefs about general teaching strategies may span many academic domains. However, given the multidimensionality and specificity of beliefs about knowledge in different academic domains, we speculated about the via­bility of applying similar epistemic belief perspectives to a professional knowledge domain, specifically teaching knowledge.

Some studies of teachers’ epistemic beliefs appear to address beliefs about the nature of teaching knowledge. For instance, Brownlee et al. (2001) examined preservice teachers’ beliefs in the context of a teacher education course through interviews and structured journals. White (2000) interviewed preservice teachers about how they would respond to “problematic classroom situations” (p. 281) in order to ascertain beliefs about the certainty, simplicity, and source of knowledge as well as their justification for knowing. Using a constant comparative data analysis method, participants were classified into categories that were somewhat reflective of King and Kitchener’s (1994) model of reflective judgment. That is, White (2000) categorized her preservice teachers as departing absolutist, intuitive relative, selective relative, informed rela­tive, and reflective relative. This work highlights the variance in teach­ers’ beliefs as well as how their beliefs may influence how they approach classroom instruction.

Teacher knowledge

As we embarked on the study of teachers’ beliefs about teaching knowl­edge, we were cognizant of the need to define what is meant by teach-ing knowledge. Here we turned to the literature on teacher knowledge. Within this literature, numerous taxonomies or frameworks have attempted to classify the knowledge that should be a part of a teach­er’s knowledge base (e.g., Carter, 1990; Elbaz, 1983; Grossman, 1990;

Helenrose Fives and Michelle M. Buehl 475

Leinhardt and Smith, 1985; Shulman, 1987). For instance, based on a case study of a high school English teacher, Elbaz (1983) identified five domains of teachers’ practical knowledge: knowledge of self, the milieu of teaching, subject matter, curriculum development, and instruction. Shulman’s (1987) categorization scheme of the teacher knowledge base is a commonly cited framework for understanding the compo­nents of teachers’ knowledge. He organized teachers’ knowledge into seven areas: content knowledge (i.e., subject area to be taught such as mathematics or science), general pedagogical knowledge (i.e., content­free teaching strategies applied across multiple content areas, such as classroom management or general instructional strategies), curricu­lum knowledge (i.e., knowledge of the scope and sequencing of content areas from the beginning of schooling through its completion), peda­gogical content knowledge (i.e., content­specific teaching strategies), knowledge of learners and their characteristics (i.e., awareness of learn­ers’ development, culture, and personal characteristics), knowledge of educational contexts (e.g., knowledge of schools, classrooms, or muse­ums), and knowledge of educational ends, purposes, and values, and their philosophical and historical roots.

Other knowledge taxonomies have also been proposed (e.g., Munby et al., 2001; Holt­Reynolds, 2000; Jordan and Stanovich, 2003; Tsai, 2002), but, unfortunately, consensus had not been reached, nor have terms been clearly defined. While issues regarding What is teaching know-ledge? will likely continue to plague teacher education, teacher assess­ment, and teacher practice, we felt that it is important to explore what knowledge teachers, not just researchers or teacher educators, believe is the necessary knowledge for teaching. Further, we speculated that given the domain­specificity of knowledge beliefs as well as the complexity of the knowledge needed for teaching, individuals may hold different beliefs about the different aspects of teaching knowledge. (Note: we use aspects of knowledge to refer to the different bodies of knowledge indi­viduals may refer to in their responses.)

Purpose and rationale

Thus, the purpose of the work presented here is to explore how pre­service and practicing teachers conceptualize the knowledge they need for teaching and their beliefs about the nature of that knowledge, as well as beliefs about the origins of the ability to teach. Identifying such beliefs provides an opportunity to understand teacher cogni­tion more fully and to examine these beliefs in relation to teacher

Teachers’ articulation of beliefs about teaching476

education and practice. Additionally, we believe that teachers’ beliefs about teaching knowledge may inf luence how and what they gain from teacher education and professional development opportunities. For instance, we speculated that beliefs about the stability or source of knowledge may inf luence teachers’ responses and receptivity to professional development opportunities. Beliefs about the structure of teaching knowledge may affect the extent to which new informa­tion is elaborated on and connected to prior knowledge. Beliefs about the ability to teach may determine how master teachers, cooperating teachers, and administrators respond to student teachers and strug­gling new teachers.

In our review of the literature, we did not identify any measures that adequately assessed individuals’ beliefs about the aspects of teaching knowledge that we wanted to explore. To address this apparent gap, we borrowed from grounded theory (Bogden and Biklen, 1992; Glaser and Strauss, 1967) and asked preservice and practicing teachers what they believed about teaching knowledge in an effort to understand how these beliefs manifest and are articulated by practitioners.

Methodology

Because our goal was to identify the maximum amount of diversity in how preservice and practicing teachers described their beliefs about teaching knowledge and to identify emergent themes that represented the full spectrum of the beliefs articulated, we were not interested in quantifying the frequency with which beliefs were reported, nor were we interested in comparing preservice and practicing teachers. The former of these possible purposes would not lead to a broad identifi­cation of possible beliefs, which is our goal. The latter of these poten­tial purposes, although important and interesting, would be better addressed through research methods designed to examine these dif­ferences. For instance, a large scale quantitative study could examine subgroups of preservice (e.g., before student teaching, student teach­ers, etc.) and practicing teachers (e.g., early career, mid­career, etc.) in relation to their beliefs about teaching knowledge. Alternatively, an in­depth comparative case study that employs purposeful sampling, observations, and interviews could offer descriptive and explanatory comparisons for teachers at various stages in their career develop ment. However, the methods we used and data gathered do not address these issues. Thus, herein we do not emphasize the frequency of reported beliefs or seek to make comparisons between preservice and practic­ing teachers.

Helenrose Fives and Michelle M. Buehl 477

Participants

To obtain a variety of perspectives, we sampled both preservice and practicing teachers from two state universities in the Southwest and Midsouth regions of the United States. Our sample consisted of fifty­three preservice and fifty­seven practicing teachers who were primarily female (preservice: 70 percent; practicing: 79 percent) and European American (preservice: 70 percent; practicing: 80 percent). Participants were teaching or planned to teach at the elementary (i.e., preservice: 34 percent; practicing: 35 percent), middle school (i.e., preservice: 6 percent; practicing: 28 percent), and high school levels (i.e., pre­service: 28 percent; practicing: 16 percent) in a variety of content areas.

Questionnaire

To uncover these preservice and practicing teachers’ beliefs about teaching knowledge, we developed ten open­ended and two restricted response questions (see Fives and Buehl, 2008, for a copy of all the questions). The results synthesized here are drawn from our analysis of a subset of these items (i.e., items 2, 4, 7, 8, and 9). Through our analysis process we found that participants’ responses to these items provided the most relevant information regarding their beliefs about the source (item 7), stability (items 6a and 6b), and content of teaching knowledge (items 4 and 8), as well as beliefs about ability to learn to teach (items 2 and 9; see Table 15.1).

Analysis procedures

Written responses to all questions were transcribed into a spreadsheet and analyzed from a modified grounded theory perspective (Glaser and Strauss, 1967), employing the constant comparative method of data analysis (e.g., Bogdan and Biklen, 1992; Strauss and Corbin, 1998). Grounded theory is an inductive approach to research in which theory is developed from data rather than applying theory to data. We consider our work to be a modified version of grounded theory research because we did not begin our research design or data analysis as theoretical. However, we also did not set out to prove an existing theory.

Through the initial transcription process we were able to identify specific items that yielded responses that informed us about our par­ticipants’ beliefs about teaching knowledge. We analyzed the specific items that addressed the research questions we sought to answer (see Table 15.1). Specifically, we coded items 4 and 8 together to address

Teachers’ articulation of beliefs about teaching478

the question of what is teaching knowledge, items 2 and 9 were coded together to examine beliefs about teaching ability, items 6a and 6b allowed us to understand beliefs about the stability of teaching knowl­edge, and item 7 provided us with information about participants beliefs about the source of teaching knowledge.

All data were initially coded at the concept level using the con­stant comparative method (e.g., Bogdan and Biklen, 1992; Strauss and Corbin, 1998). That is, each complete concept or idea unit was assigned a code that reflected the main idea of the concept. The data set was divided into three lots; we both coded the first lot independ­ently, compared the code generated, and together developed an initial code sheet. This code sheet was applied by one of the authors to the second lot of data; as new concepts were encountered new codes were generated and added to the code sheet. The other author then used the extended code sheet to analyze the third lot of data, added codes as necessary, then re­coded the first two lots with the more elaborate code sheet. Finally, the original coder used the elaborate codes sheet to re­code all of the data. We then compared our coding of the data. Instances of disagreement were discussed and final codes were agreed upon by both authors.

Following this fine­grained coding of the data, we examined these codes for emergent themes by independently engaging in a physical sort of the generated codes related to each question set. Using ana­lytic induction, we independently grouped similar codes into themes and developed labels for those themes (Strauss and Corbin, 1998). We then compared the themes we identified independently. All like themes were retained. Any differences were discussed (and debated) until we reached consensus as to the best way to represent the data.

Teaching knowledge belief framework

The teaching knowledge belief framework is comprised of five frames (i.e., domain knowledge, source of knowledge, ability beliefs, stability of knowledge, domain qualities and skills) that emerged from our data and reflect the preservice and practicing teachers’ responses to questions about teaching knowledge. The specific analyses and findings leading to these frames are discussed in detail in our previous empirical papers (i.e., Buehl and Fives, 2009; Fives and Buehl, 2005; Fives and Buehl, 2008). Here, for the first time, we present a synthesis of this work in an attempt to articulate our framework to this point. For each frame, we describe the beliefs that compose the frame and discuss their implica­tions for teacher education and practice.

Helenrose Fives and Michelle M. Buehl 479

Table 15.1 provides an overview of our framework. Each frame is comprised of emergent themes that reflect the multiple perspectives our participants articulated. To highlight the diversity and coherence of one teacher’s beliefs, we begin the discussion of each frame with a quotation from the same practicing second grade teacher of five years.

Domain knowledge: what is teaching knowledge?

Effective teachers understand child development, classroom management, and have content knowledge. They need to understand what type of instruction[al] methods could be used to teach certain things and where curriculum can be found that would enhance instruction. (ID 421, practicing teacher)

We refer to our first frame as domain knowledge wherein we conceptual­ize the field of teaching as a domain. Alexander (1992) defined domain knowledge as “the realm of knowledge that individuals share about a particular field of study” (p. 34). According to Alexander (1992), this realm of knowledge includes knowledge that is declarative, procedural, or conditional in nature (Paris et al., 1983). We adopt this perspective and offer teachers’ knowledge of teaching as a domain. In particular, we recognize this domain of knowledge as one for which practition­ers would have specific beliefs that may influence their perspectives on teaching, learners, and their practice.

When asked “What knowledge is necessary for effective teaching?” and “What knowledge do teachers hold that is unique to the teach­ing profession?” participants offered a myriad of responses which we felt represented five distinct themes (see Table 15.1). These themes included classroom management and organization, pedagogical knowledge, content knowledge, children, and self and other fields. These themes are reflective of the knowledge categorization schemes proposed and dis­cussed by researchers (e.g., Elbaz, 1983; Richardson, 1996; Schulman, 1987).

Classroom management and organization. Participants described the need for teachers to know about classroom management related to stu­dent behavior or classroom control as well as the more administrative and organizational aspects of the teaching profession. The responses organized within the classroom management theme indicated that classroom management is more than just maintaining student behav­ior. The need for teachers to “multi­task” (ID 165, practicing) and to have “time­management” (ID 171, practicing) were seen as impor­tant forms of knowledge. Interestingly, the majority of the comments describing the need for teachers to multi­task came from practicing

Tab

le 1

5.1.

Fra

mew

ork

of b

elie

fs a

bout

teac

hing

kno

wle

dge

Org

aniz

ing

fram

ewor

k of

tea

chin

g kn

owle

dge

bel

iefs

Que

stio

n

Item

s

Wha

t is

tea

chin

g kn

owle

dge?

Whe

re d

oes

teac

hing

kn

owle

dge

com

e fr

om?

Whe

re d

oes

teac

hing

ab

ility

com

e fr

om?

Doe

s te

achi

ng k

now

ledg

e ch

ange

?W

hat

qual

itie

s or

ski

lls d

o te

ache

rs n

eed

to h

ave?

(4) W

hat

know

ledg

e is

ne

cess

ary

for

effe

ctiv

e te

achi

ng?

Ple

ase

be

spec

ific.

(2)

Is t

each

ing

a ta

lent

pe

ople

are

bor

n w

ith?

P

leas

e ex

plai

n.

(2)

Is t

each

ing

a ta

lent

peo

ple

are

born

wit

h? P

leas

e ex

plai

n.

In t

he n

ext

twen

ty

year

s …

(6

a) h

ow m

uch

do y

ou

thin

k th

e kn

owle

dge

need

ed f

or e

ffec

tive

te

achi

ng w

ill c

hang

e?

(4) W

hat

know

ledg

e is

ne

cess

ary

for

effe

ctiv

e te

achi

ng?

Ple

ase

be s

peci

fic.

(8) W

hat

know

ledg

e do

tea

cher

s ho

ld

that

is u

niqu

e to

the

te

achi

ng p

rofe

ssio

n?

(7) W

here

doe

s kn

owle

dge

of h

ow t

o te

ach

com

e fr

om?

(9)

Can

som

eone

le

arn

how

to

be a

n ef

fect

ive

teac

her?

P

leas

e ex

plai

n.

(6b)

in w

hat

way

(s)

do y

ou t

hink

the

kn

owle

dge

need

ed f

or

teac

hing

will

cha

nge?

P

leas

e pr

ovid

e sp

ecifi

c ex

ampl

es.

(8) W

hat

know

ledg

e do

te

ache

rs h

old

that

is u

niqu

e to

th

e te

achi

ng p

rofe

ssio

n?

(9)

Can

som

eone

lear

n ho

w t

o be

an

effe

ctiv

e te

ache

r? P

leas

e ex

plai

n.

(2)

Is t

each

ing

a ta

lent

peo

ple

are

born

wit

h? P

leas

e ex

plai

n.

(9)

Can

som

eone

lear

n ho

w t

o be

an

effe

ctiv

e te

ache

r? P

leas

e ex

plai

n.

Org

aniz

ing

fram

ewor

k of

tea

chin

g kn

owle

dge

bel

iefs

Fra

me

Dom

ain

know

ledg

eS

ourc

e of

kno

wle

dge

Abi

lity

belie

fsS

tabi

lity

of k

now

ledg

eD

omai

n qu

alit

ies

and

skill

s

Em

erge

nt

The

mes

Cla

ssro

om

man

agem

ent

and

orga

niza

tion

For

mal

pre

para

tion

Inna

teA

mou

nt o

f ch

ange

Com

mun

icat

ion

Ped

agog

ical

kn

owle

dge

For

mal

ized

bod

ies

of

info

rmat

ion

Req

uire

s po

lish

Dir

ecti

on o

f ch

ange

Ada

ptab

ility

and

inge

nuit

y

Con

tent

kno

wle

dge

Inte

ract

ive

and

colla

bora

tive

exp

erie

nces

Inna

te a

nd le

arne

dQ

ualit

y of

cha

nge

Nur

tura

nce

and

care

Chi

ldre

nO

bser

vati

onal

and

vi

cari

ous

expe

rien

ces

Lea

rned

Spe

cific

top

ics

– w

hat

chan

ged?

Ent

husi

asm

Sel

f an

d ot

her

field

sE

nact

ive

expe

rien

ces

Cal

ling

or g

ift

Rea

sons

for

cha

nge

Inte

grit

y an

d co

mm

itm

ent

Sel

f­re

flect

ion

Teachers’ articulation of beliefs about teaching482

teachers. Thus, it may be that preservice teachers are unclear of the depth of the administrative aspects of the teaching profession. Per­haps this is an area of knowledge that deserves greater attention within teacher education experiences.

Pedagogical knowledge. We describe the second theme of this frame as pedagogical knowledge; in doing so we refer to those aspects of teaching that are reflective of the professional knowledge and skills needed for instructional activities. Thus, included in this theme are facets reflecting: knowledge about teaching methods and practices, assessment, motivation, and reaching students. For instance, knowl­edge about teaching methods and practices included both declara­tive and procedural knowledge. One preservice teacher described this knowledge as

effective classroom practices; … [teachers] know how to make knowledge accessible to students; … conducting hands­on learning activities; knowledge about managing and mentoring student learning. (ID 405, preservice)

Content knowledge. Knowledge of content emerged as a theme within the frame of domain knowledge. This theme includes knowledge of the specific material to be taught, content­specific pedagogy, and cur­riculum. Expertise in the content area was considered essential as evi­denced by one practicing teacher who stated “content knowledge – need to be an expert in what you are teaching” (ID 125). However, what also emerged was a need to know the content specific methods for how that unique content should be taught for specific groups of learners (i.e., ped­agogical content knowledge; Shulman, 1987) and how the content fits into the overall curriculum. For example, one preservice teacher stated that teachers needed “[e]xcellent knowledge of the subject area, … how the subject should be presented and assessed” (ID 401). This response illustrates the need to not only be an expert in the content, but also be aware of and able to employ content­specific pedagogy.

Knowledge of children. Awareness of learners in general and the stu­dents in their own classes emerged as critical facets of knowledge nec­essary for teaching. At a general level, participants described a need for teachers to be aware of such topics as child development, learning disorders, how children learn, the environmental influences on learn­ing (i.e., parents, socio­economic status, culture), and individual differ­ences. In addition to these more general areas, knowledge of the specific students in their own classroom was also seen as critical knowledge for teachers’ knowledge base. Respondents reported that it was imperative for teachers to know their students’ strengths, weaknesses, family, cul­tural background, and individual needs. A practicing teacher described

Helenrose Fives and Michelle M. Buehl 483

the need for these general and specific aspects related to knowledge of children in her response:

Background knowledge of child development (education classes); … individual needs of students, including developmental level; background/culture/home life of students – understand where they’re coming from. (ID 125)

Knowledge of self and other fields. The final theme that emerged related to the frame of domain knowledge pertained to knowledge of self and other fields. The need for self awareness of identity, strengths, and weaknesses were reported as necessary for effective teaching. In partic­ular, one preservice teacher stated “Knowledge of self; know your own weaknesses and don’t hold students’ accountable for them” (ID 177), highlighting the need for self­awareness when working with others.

Participants also reported that teachers should have knowledge of a variety of other topics, including pop culture, current events, soci­ety, “life knowledge,” and how to do research. Essentially, there was a perceived need for teachers to be aware of the world beyond their own classroom and content area.

Implications of conceptualizations of domain knowledge. Although the five knowledge themes we identified are reflective of knowledge catego­ries identified by others, such as Elbaz (1983) and Shulman (1987), it is important to note that our participants (1) typically identified more than one theme of knowledge in their response and (2) did not identify all types of knowledge in their responses. Thus, participants recognized that within the domain of teaching knowledge there are multiple bodies of information that must be harnessed for successful or effective teach­ing to occur. However, not all participants viewed all of these bodies of information as necessary for teaching. Future research may benefit from examining the valuing of these bodies of knowledge in more detail. For instance, studies could be designed to explore differences based on level of experience (e.g., preservice, novice, or experienced teachers) as well as grade level (e.g., elementary, middle school, and high school). Given that we had unequal numbers of preservice and practicing teachers at each grade level and that participants were not probed to provide or rank all possible teaching knowledge, we did not feel we could address these issues with our current data.

Researchers interested in understanding preservice and practicing teachers’ beliefs about teaching knowledge must address the complex­ity of this field and recognize that individuals hold complex, multifac­eted perspectives on this domain. If a general measure of beliefs about teaching was to be developed it might miss some of the nuanced beliefs that are held by teaching professionals. When researchers ask teachers

Teachers’ articulation of beliefs about teaching484

about teaching knowledge, the particular aspect of teaching knowledge should also be specified. Others in the field have also recognized the need for specificity. For example, the teacher sense of efficacy scale (TSES) developed by Tschannen­Moran and Woolfolk­Hoy (2001), specifically addresses efficacy beliefs relative to instructional practices, classroom management, and student engagement. Our research suggests that such specificity in Likert scale measures is appropriate and, depending on one’s purposes, even greater specificity may be warranted. Similarly, in interviews and case studies, researchers need to be explicit about the aspects of teaching knowledge beliefs they are exploring. For instance, although White (2000) may have addressed preservice teachers’ beliefs about teaching knowledge through her interviews about problematic classroom situations, the nature of the imagined scenario may have influenced the types of beliefs that were elicited from the participants (e.g., classroom management versus content knowledge).

Teacher educators and those interested in providing professional development opportunities for practicing teachers must also acknowl­edge the varied perspectives preservice and practicing teachers hold with respect to different aspects of teaching knowledge and take this into account in their instruction. Indeed, Patrick and Pintrich (2001) noted the role of teachers’ beliefs and motivation in teacher conceptual change and underscored the need for teacher educators to be aware of preservice teachers’ beliefs. For instance, they suggested that preservice teachers’ adoption of mastery goals were likely to lead to a deeper level of engagement and processing in teacher education classes and that task value has been related to greater use of cognitive and meta­cognitive strategies (Pintrich and Schrauben, 1992). In contrast, performance goals may induce preservice teachers to work for a good grade but not lead to meaningful conceptual change when the new information con­flicts with prior conceptions about teaching and learning (Patrick and Pintrich, 2001).

From our perspective, preservice teachers’ adoption of mastery or performance goals as well as their valuing of tasks within teacher edu­cation classes may be related to the importance they place upon differ­ent aspects of the knowledge needed for teaching. Specifically, not all bodies of knowledge were identified by all participants as “necessary for effective teaching” or “unique to the teaching profession.” Depend­ing on the perspectives of the audience one is trying to reach, teacher educators may encounter varying degrees of acceptance and resistance. For example, future and practicing teachers interested in working with young children in elementary schools may feel that knowledge of child development is crucial to their practice as teachers, while expert subject

Helenrose Fives and Michelle M. Buehl 485

matter knowledge may seem less important to these professionals. The reverse of this may be true for secondary teachers. Similarly, preservice and practicing teachers in performance areas such as art, music, and physical education may perceive knowledge about traditional assess­ments, literacy, and motivation as unimportant to their professional practice as they do not use traditional assessments, assign reading, or work with students compelled to be in their classes.

Consequently, when preservice teachers are exposed to bodies of knowledge they view as important to teaching (e.g., instructional strat­egies), they may be more likely to adopt mastery goals, whereas if the knowledge is viewed as more tangential or even irrelevant (e.g., theories of learning and cognition), performance goals may be adopted. Further, teacher candidates are not likely to engage deeply with knowledge they do not recognize or value as necessary for teaching. Thus, they may overlook the connections of some aspects of knowledge to their own practice and miss out on important learning opportunities. As noted by Holt­Reynolds (2000), preservice teachers’ perceptions of the value of practical strategies and techniques and their devaluing of a deep under­standing of theory may not serve them well.

Further, practicing teachers are noted as being particularly resistant to new models and ideas about teaching and learning (Richardson and Placier, 2001). Such resistance may be based, in part, on practicing teachers’ beliefs about the value of certain aspects of teaching knowl­edge. Other factors that may play a role are beliefs about the certainty of teaching knowledge, which we will address in a subsequent section.

Source of knowledge: where does teaching knowledge come from?

[Individuals learn to teach] through experience and support from adminis­trators and other staff members. Having many opportunities to attend pro­fessional development for content, class management, special populations, or anything supported by research as a best practice. Observing master teachers at work. (ID 421, practicing teacher)

The second frame pertains to beliefs about the source of teach­ing knowledge. Epistemic belief research has identified the source of knowledge as a dimension of individuals’ knowledge beliefs (e.g., Jehng et al., 1993; Hofer, 2000; Schraw et al., 2002). Previous items intended to assess beliefs about the source of knowledge focus on knowledge coming from an authority figure (e.g., expert, professor, or textbook) or being developed through personal experience and reason (e.g., Bendixen

Teachers’ articulation of beliefs about teaching486

et al., 1998; Hofer, 2000). In his discussion of teacher knowledge, Shulman (1987) identified four sources of pedagogical knowledge (i.e., “[1] scholarship in the content disciplines, [2] the materials and settings of the institutionalized educational process, [3] research on schooling, social organizations, human learning, teaching and development, and [4] the wisdom of practice itself,” p. 8). Richardson (1996) articulated three categories of experience that influence teachers’ knowledge and beliefs: personal experiences, experience with schooling and instruc­tion, and experience with formal knowledge.

Our analysis of preservice and practicing teachers’ responses revealed that individuals view teaching knowledge as coming from a variety of sources. We organized these sources into six themes: formal preparation, formalized bodies of knowledge, observational and vicarious experiences, interactive and collaborative experiences, enactive experiences, and self reflec-tion. Some of the sources are external to the individual, representa­tive of knowledge coming from an authority outside of oneself, whereas other sources are reflective of experience and self­reflection as sources of knowledge. It is also notable that the sources identified by our pre­service and practicing teachers are reflective of the sources of knowl­edge identified in the personal epistemology (e.g., Bendixen et al., 1998; Hofer, 2004) and teacher knowledge (e.g., Elbaz, 1983; Richardson, 1996; Shulman, 1987) literatures.

Formal preparation. Within this frame, two themes reflected external authority as the source of teaching knowledge. Specifically, individuals referred to the importance of teacher education classes and professional development experiences as salient sources of knowledge for teach­ing. For instance, several preservice teachers acknowledged the impor­tance of college coursework, stating “we get some knowledge of how to teach from long hours of college work” (ID 155), and “ Hopefully college otherwise why am I here? College will help me with facts” (ID 111). While both preservice and practicing teachers acknowledge formal preparation as a source of teaching knowledge, this was never the only source mentioned by our respondents. Most participants who reported formal preparation as a source also indicated that col­lege courses were insufficient on their own and that more experiential sources of knowledge were also needed (these are described in three of the themes below).

Formalized bodies of information. This theme reflected an external authority as the source of teaching knowledge. This theme included ref­erences to particular information stores (e.g., books, articles) and accu­mulated findings from research. For instance, a practicing elementary teacher indicated that teaching knowledge comes from “professional

Helenrose Fives and Michelle M. Buehl 487

literature, conferences, course work, studies” (ID 408), whereas another practicing teacher indicated that “[t]his knowledge comes from so many fields: psychology, medicine, neurology, social science, child development, and nutritional experts” (ID 235).

Observational and vicarious experiences. Three themes identified the source of teaching knowledge as various types of personal experiences. Observational and vicarious experiences (e.g., “watching model teach­ers,” ID 277, practicing) were cited as important sources for knowledge of how to teach. The observational experiences reported included both naturally occurring observations such as “You are taught all through your life by observational learning” (ID 101, preservice), and more planned observations such as “How to teach comes from … modeling during student teaching and field observation” (ID 113, preservice).

Interactive and collaborative experiences. Another theme that empha­sized the experiential nature of the source of teaching knowledge included responses that described interactive and collaborative experi­ences with others (e.g., “collaborating with experienced teachers,” ID 424, practicing). Both preservice and practicing teachers reported that teaching knowledge came out of collaborative experiences with other teachers.

Enactive experiences. Participants also frequently cited enactive experiences as a source of teaching knowledge. Both practicing (e.g., “mostly by getting in there and teaching – I learned best just doing it,” ID 411, practicing) and preservice (e.g., “The knowledge of how to teach comes from practice and experience,” ID 124, preservice) teach­ers’ recognized direct experiences with learners as a source of teaching knowledge.

Self-reflection. We identified a final theme related to the self as a source of teaching knowledge. Specifically, responses indicated that teach­ing knowledge came from within oneself. Reflection on one’s experi­ences as a source of knowledge (e.g., “reflecting on own experience in school and with teachers,” ID 207, preservice), and one’s weaknesses and shortcomings (e.g., “examining your shortcomings and trying to improve your faults when dealing with youths,” ID 253, preservice), were represented as two sources of teaching knowledge.

A synthesis of personal learning was also identified as a source of teaching knowledge. One preservice elementary teacher indicated that “The knowledge of how to teach comes from experience. It comes from synthesizing everything you’ve learned” (ID 117). These facets of the self as a source of knowledge, as well as the themes related to experience as a source of knowledge, are reflective of beliefs previously espoused in the literature.

Teachers’ articulation of beliefs about teaching488

Another facet of the self as a source of knowledge represented a belief in innate knowledge of how to teach. That is, respondents indicated that individuals were born with knowledge of how to teach. For instance, when asked where the knowledge of how to teach comes from, a pre­service teacher responded:

It is a talent, it comes from within. You must have head knowledge and back­ground information to teach, but knowing how to teach comes from inside. You are born with it. (ID 120)

The issue of innate ability is addressed in the next section on beliefs about the ability to teach. However, we found it intriguing that indi­viduals also expressed a belief in innate knowledge.

Implications for articulated beliefs about the sources of knowledge. Our examination of individuals’ beliefs about the sources of teaching knowl­edge suggested several implications. In particular, we were struck by the number of sources individuals listed. Rarely did a respondent indi­cate a single source of teaching knowledge. Our participants, instead, typically listed several sources of knowledge and identified the sources of knowledge in concrete terms. For example, one practicing teacher responded that

[teaching knowledge] is learned from – (1) Watching humans interacting with each other; (2) studying content areas (college); (3) making mistakes and learn­ing from them; (4) practicing in a classroom. (ID 423)

Responses also suggested that sources of knowledge may be weighted or valued differently, such that some sources of teaching knowledge may be perceived as more or less important to a teacher’s construction of meaning about teaching practice. For instance, one preservice teacher noted that her knowledge of how to teach came from “[p]ersonal expe­riences mostly. I have learned some things about education in my classes but I have learned more from being in classes during my observations” (ID 104; emphasis added).

Both the number and differential valuing of sources have implica­tions for research and teacher education. Specifically, a single fac­tor or continuum may not adequately represent individuals’ beliefs about the sources of teaching knowledge. Questionnaires developed to assess such beliefs need to offer a means for multiple sources to be selected as well as a means for individuals to indicate the relative con­tribution of each source. Further, one practicing teacher stated that “[o]f course, you learn teaching strategies, pedagogy, or even teaching methods in your classes, but real knowledge come from actual prac­tice” (ID 415), suggesting that there may be different beliefs about

Helenrose Fives and Michelle M. Buehl 489

the source of teaching knowledge depending on the aspect of teaching knowledge under consideration. Perhaps it would be useful to examine individuals’ beliefs about the sources of the different bodies of teach­ing knowledge.

Implications for teacher education are also evident in our participants’ responses. The differential weighting of the various sources of knowl­edge reflected distrust or questioning of formal education and formal­ized bodies of knowledge as legitimate sources of teaching knowledge. One practicing teacher reported that “[u]nfortunately I feel the theories come from professors that really have no idea about kids from all learn­ing levels or socio­income areas. Things that we learned over 20 years ago were not realistic in the city schools” (ID 265).

Such beliefs are concerning because they may influence how preserv­ice and practicing teachers respond to formal educational experiences including the coursework required for licensure and continued certi­fication as well as professional development opportunities. If teachers devalue formal education and formal bodies of information as legiti­mate sources of teaching knowledge, they may not actively pursue addi­tional education nor consume more recent research related to methods of teaching and learning. Further, when teachers with such beliefs are required to participate in professional development experiences, they may be less attentive and less likely to consider the new information they are presented (e.g., Book et al., 1983; Holt­Reynolds, 2000; Patrick and Pintrich, 2001; Richardson, 1996).

Within the personal epistemology literature, viewing knowledge as coming less from authority and more from one’s own experience and active construction of meaning is considered to be a more sophisti­cated stance (e.g., Hofer and Pintrich, 1997; Hofer, 2004). This would imply that those teachers who place more value on personal experience as a source of knowledge than they do on formal education would have a more sophisticated perspective. However, one must also consider the justification of knowledge. That is, one’s knowledge may be evaluated based on various grounds including what an authority indicates is cor­rect, one’s own feeling or perception of what is correct, or a critical and reasoned evaluation of the available claims and evidence (e.g., Hofer, 2004; King and Kitchener, 1994). Thus, perhaps teacher education and professional development should promote preservice and practic­ing teachers’ critical evaluation of what they hold to be true relative to teaching based on the analysis of evidence and theories provided by formal education and formalized bodies of teaching knowledge, as well as taking into account their consideration of alternative views (e.g., Patrick and Pintrich, 2001).

Teachers’ articulation of beliefs about teaching490

Ability beliefs: where does teaching ability come from?

No, teaching is not a talent people are born with. I believe teachers through experience in a classroom are able to hone skills and become master teachers. It is through experience, staff development and a personal commitment to excel­lence that people become teachers. (ID 421, practicing teacher)

Schommer (1990), now Schommer­Aikins, described a hypothesized model of beliefs that included beliefs about the stability, source, and structure of knowledge, as well as beliefs about the speed of learn­ing and the ability to learn. In 2004, Schommer­Aikins updated her organization and offered an embedded systemic model in which she separated beliefs about knowledge (i.e., stability, source, and structure) from beliefs about learning (i.e., speed of learning and ability to learn). We similarly recognize the importance of ability beliefs as an essential component in the larger belief system. Moreover, we were interested in beliefs about the ability to learn how to teach. Students’ beliefs about ability have received considerable attention (e.g., Dweck, 2002; Sti­pek and MacIver, 1989). In our view, beliefs about the ability to teach may influence preservice and practicing teachers’ learning processes, approaches to coursework and professional development, and interac­tions with mentors or protégés.

Participants’ responses to three questions (Table 15.1) tapping into their beliefs about the ability to learn to teach were analyzed and five themes emerged. The emergent themes described the ability to teach as (1) innate, (2) requires polish, (3) innate and learned, (4) learned, and (5) a calling or gift.

Teaching ability is innate. The perspective that teaching ability is innate was articulated in a variety of ways. Participants reflecting an innate perspective described teaching ability as something one is born with, others called it an art or a “knack” that comes from within, a tal­ent, instinct, and based on one’s personality. For example a practicing teacher stated:

I feel teaching is an innate talent. Many people “know” a subject, but to be able to transfer that information in various ways in order for another to learn and internalize that information and access it later requires a true talent. (ID 415)

This statement highlights this teacher’s distinction between the ability to teach versus the ability to “know” a particular subject or academic domain. This statement underscores the need to examine teaching as a distinct domain about which beliefs are held.

Helenrose Fives and Michelle M. Buehl 491

Teaching ability requires polish. Most participants were more tenta­tive in their beliefs about the ability to teach and argued for a more nuanced perspective. For example, many felt that although some aspects of teaching ability are innate, raw talent is not enough. Training is required for that talent to become useful in the classroom. A practic­ing high school teacher of twenty­one years exemplified this perspec­tive, stating: “A ‘seat of your pants’ feel for how to reach others and communicate easily is inborn, but effective methods and strategies can be learned” (ID 406).

Teaching ability is innate for some, but learned for others. Another more nuanced perspective on the issue of innate versus learned teaching abil­ities was that some participants indicated that for some people teaching ability was innate, but others can learn how to teach. That is, some people seem to have a natural talent or tendency for teaching and others require extensive training and education to do so. Within this perspec­tive, however, was the belief that all people can learn to teach, and that depending on one’s talent the learning experiences may be more or less difficult.

Teaching ability is learned. In contrast to the perspectives described above, many felt that the ability to teach is learned; as one preserv­ice teacher stated “no one is born with a ‘teaching gene.’ Training is needed” (ID 114). Other responses within this perspective focused on the idea that the ability to teach can be acquired through education and training. An elementary school teacher of sixteen years articulated a learned belief about teaching ability, with the caveat that certain cir­cumstances needed to be met for that learning to occur.

You can learn how to be an effective teacher. You need (a) the desire to improve; (b) adequate instruction (classroom, workshops, direct observation, discussion groups, reading); (c) adequate support (mentor, administration, curriculum); (d) adequate environment (materials, classroom, proper class size). (ID 245)

Thus, while many described a belief that teaching can be learned, they recognized that this learning was not easily done and required a variety of scaffolds from the learning environment.

Teaching is a calling or gift. Theoretically distinct from an innate abil­ity, a talent requiring polish, or a skill to be learned, this last set of beliefs, teaching as a calling or gift, emphasized the ability to teach as not from within nor from instruction but as an ability bestowed by a greater being or higher source. Frequently, responses in this theme reflected a “you have it or you don’t” perspective similar to an innate perspective on teaching ability. In our analysis, we felt that this theme

Teachers’ articulation of beliefs about teaching492

might be distinct from an innate perspective because the source of abil­ity comes from outside the individual as a gift or calling from a higher being. For example:

Teaching is an art and a gift from God. Some are blessed with the ability to be effective and some aren’t. Anyone can stand in a class and spit out a bunch of knowledge but to be an effective teacher you must have that gift. (ID 101, preservice)

Others emphasized that teaching is a calling, and described the profes­sion as a vocation they were meant to do. Connected to the belief that teaching is a calling was the notion that those who were not called to teach would not remain in the profession. This sentiment was voiced by experienced and preservice teachers.

I truly believe teaching is a calling that you either have it or not. You cannot try to become a teacher. Those who do quickly leave the profession. (ID 413, practicing)

to teach is a calling to me. I knew that ever since I was a little girl that I was going to be a teacher. I feel it in my bones! There is nothing else that I’d rather be. Some people teach but they are not called to teach, that’s why they don’t stick with it but if you are born a teacher, you’ll retire a teacher. (ID 109, practicing)

Implications of ability beliefs. The identified variations in our participants’ beliefs about the ability to teach have implications for the assessment of ability beliefs and offer insight into issues related to teacher devel­opment, retention, and practice. Specifically, the nuanced gradations of beliefs suggest the need for more sensitive measures that can cap­ture the complexity of these beliefs. In particular, individuals with more mixed or hybrid views of teaching ability pose a particular challenge in that they may similarly endorse items that reflect innate (i.e., entity) and acquired (i.e., incremental) ability beliefs (Dweck, 2002).

The majority of research that has looked at the influence of abil­ity beliefs has done so with school age and undergraduate popula­tions relative to typical school subjects (e.g., Henderson and Dweck, 1990; Hong and Dweck, 1992). Moreover, these studies have found that learners holding entity beliefs (i.e., view attributes as unchanging, fixed, and uncontrollable) about their intelligence are likely to attribute poor performance to internal, uncontrollable causes, causes that can­not be altered and are seen as the fault of the learner (e.g., Henderson and Dweck, 1990; Hong and Dweck, 1992). In contrast, learners hold­ing incremental beliefs (i.e., view attributes as dynamic and malleable) attribute poor performance to low effort or factors related to the task.

Helenrose Fives and Michelle M. Buehl 493

Research has yet to establish if the more hybrid views of ability described by some of our participants will lead to similar findings or if these hybrid beliefs can be used to foster adaptive attributions in teachers. It may be the teachers who believe teaching requires an innate talent that needs to be honed through formal education, professional practice, and reflection will make more adaptive decisions when faced with failures in the classroom. This is an area for future investigation. In the conclusions section, we discuss a new measure we have devel­oped to address such issues.

We hold that beliefs about teaching ability may be a core issue for new teachers leaving the profession. This connection can be informed by an attribution theory perspective (Weiner, 1985). Beliefs about teach­ing ability as innate or a gift differ with respect to a locus of causality. Innate abilities may be viewed as having a more internal locus of causal­ity, whereas statements articulating teaching ability as a calling or gift often referred to an external source (e.g., God). However, these beliefs are similar in that individuals may view their abilities as uncontrollable and stable. Thus, individuals who believe that the ability to teach is innate or an inherent gift may experience feelings of hopelessness and low motivation when they encounter difficulties in the classroom. If they do not believe that their ability to teach can be improved, individu­als may decide to leave the teaching field before giving themselves time and opportunity to develop their abilities.

The ability beliefs of preservice and new teachers are not the only concern. Beliefs about the ability to teach may also account for vet­eran and cooperating teachers’ reactions to struggling student teachers and new teachers. Experienced teachers who believe that at least some aspects of teaching are learned are likely to provide more guidance, support, and mentoring for those entering the profession. In contrast, experienced teachers with more innate views are likely to be less sup­portive and adopt more of a sink or swim attitude to new teachers.

Studies of teachers’ beliefs about student ability support this per­spective. For instance, Butler (2000) found that teachers with differ­ent beliefs about the nature of student ability made different inferences when presented with performance data. In particular, teachers who held incremental beliefs about ability were more likely to reserve judg­ments about students’ abilities. Entity theorists were more likely to use initial information as indicative of the student’s potential. In another investigation, Lee (1996) found that not only do teachers with entity and incremental theories of ability treat students differently in terms of grading, feedback, follow­up assignments, and ability grouping, but are also more likely to be influenced by their initial expectations for

Teachers’ articulation of beliefs about teaching494

students. She concluded that teachers’ theories of intelligence may con­tribute to whether students are treated in an unbiased and appropriate or a biased and unfair way.

Research is needed to determine if such relations are evident in the relationships and interactions between novice and mentor teachers. Such research should take individual differences on the part of the mentor and novice teachers into account. For instance, mentor teach­ers many vary in their beliefs and approach to mentoring. Further, the effects of mentoring may be different if the novice and mentor teachers hold different beliefs about the ability to teach and the role of the men­tor teacher.

Additionally, beliefs about the ability to teach may be linked to the pursuit of further education and professional development at all levels of teaching experience. Individuals who believe their teaching abilities are malleable are more likely to seek out experiences to improve their skills.

Finally, we should consider the relation between beliefs about the abil­ity to teach and beliefs about the ability to learn. If teachers believe that their own ability to teach is an innate entity that cannot be improved with experience or education, do they hold similar beliefs about their students’ abilities? Such congruence in beliefs may have a detrimental impact on teachers’ expectations, motivation, and instruction within the classroom, particularly when working with students with special needs. Indeed, Jordon et al. (1997) found that teachers with more entity views of student ability provided less effective instruction for students with disabilities. However, incongruence between beliefs about the ability to learn how to teach and the ability to learn in another con­tent area would suggest that ability beliefs are domain specific, further underscoring the need to consider beliefs about teaching and teaching knowledge as a professional domain of study. In our data, there was evidence of such domain specificity.

Stability of knowledge: does teaching knowledge change?

If trends continue as they have in the last 100 years the knowledge needed for effective teaching will change quite a bit. New understandings of how people learn will emerge. Teachers will need a greater technological background, and a good knowledge of software. I’m not sure what type of technological training because this area changes rapidly. (ID 421, practicing teacher)

Within the epistemic belief literature, researchers have examined students’ beliefs about the certainty and stability of knowledge (e.g.,

Helenrose Fives and Michelle M. Buehl 495

Schommer, 1990; Schraw et al., 2002). Evidence suggests that students hold beliefs about the degree to which knowledge exists and can be attained with certainty, as well as beliefs as to whether knowledge is changing and tentative or more static and stationary. Some findings also indicate that students’ beliefs about the stability of knowledge are closely linked to their beliefs about the structure and organization of knowledge. That is, items developed to assess beliefs about the chang­ing or static nature of knowledge and items developed to assess whether knowledge is viewed as complex or more simplistic have been associated with the same underlying factor (e.g., Hofer, 2000; Qian and Alverman, 1995).

Based on our domain­specific and multidimensional approach to teachers’ beliefs about the nature of teaching knowledge, we speculated that teachers may hold different beliefs about the stability of teaching knowledge and that stability beliefs may vary depending on the body of teaching knowledge under consideration. Such conjecture was based on discussion of changes occurring in the education literature. For instance, Alexander et al. (1996) identified trends within education suggesting that some bodies of knowledge are increasing rapidly (e.g., educational technology), whereas others display more cyclical develop­ments (e.g., phonics versus whole language in literacy).

Our analysis of preservice and practicing teachers’ espoused beliefs about the stability of teaching knowledge focused on responses to two related questions: “In the next twenty years, how much do you think the knowledge needed for effective teaching will change?” and “In the next twenty years, in what way(s) do you think the knowledge needed for teaching will change? Please provide specific examples.” The responses to these questions were much more complex and nuanced than we anticipated. As detailed elsewhere (Buehl and Fives, 2009), in addition to statements about how teaching knowledge would change, participants made distinctions between anticipated changes and neces­sary, but unlikely, changes. Reasons for the changes in teaching knowl­edge were also offered in our participants’ responses (e.g., change due to changes in content knowledge or changes in society).

We identified five main themes related to the stability of teaching knowledge. Three of these themes emphasize beliefs that are more epistemic and reflect beliefs about the nature of change (i.e., amount of change, direction of change, and quality of change). Participants also offered specific topics that were more or less likely to change and reasons for change.

Amount of change. Participants discussed the changes in knowl­edge with respect to how much the knowledge would change. The

Teachers’ articulation of beliefs about teaching496

amount of change reported ranged from no change, to a little change, to a lot of change, to “drastic changes” (ID 117, preservice) in knowl­edge over the next twenty years. Thus, across our participants perspec­tives on the amount of change in the knowledge needed for teaching varied greatly.

Direction of change. Some responses indicated the direction of change. That is, while some merely specified that knowledge would change, others indicated that there would be an increase in knowledge over time. For example, “[knowledge] will increase twofold because of standards and accountability” (ID 206, preservice). Although a few references were also made to decreases, close examination of responses revealed that these participants typically referred to the requirements for teachers. One preservice teacher stated, “I think that the courses required will be decreased. I know that requirement hours for graduation have already gone down” (ID 108). We viewed such changes as reflective of changes in policies related to teaching certification more so than changes in the actual knowledge needed for teaching. A preservice teacher made this distinction evident when she stated, “I think that legally the knowl­edge for effective teaching will be able to be lower and lower, but in all the actual classroom[s] knowledge for effective teaching will always be needed” (ID 207).

Quality of change. In discussing changes in teaching knowledge, some participants referenced changes in the quality of teaching knowledge. That is, some believed that knowledge would become simplified over time whereas others believed that teaching knowledge would become more integrated and complex. A preservice teacher noted, “I know [teaching knowledge] will change, everything does. I think it will get simplified with technology” (ID 108).

References to the quality of knowledge change described the rate of change (e.g., gradual, constant, cyclical) as well as various qualitative shifts in the nature of teaching knowledge. Beliefs about such changes are evident in the response of a practicing teacher:

education seems to be cyclical as far as educational theories go. I do feel we need to become more and more knowledgeable in working with technology, and in working cooperatively. Our classrooms will become more global and we will be using resources that may not even exist right now. (ID 408)

The statement above illustrates that this teacher recognizes both itera­tive (i.e., cyclical educational theories) and incremental (i.e., technology, cooperative teaching, globalization) trends in education (e.g., Alexander et al., 1996). Moreover, this response provides insight into beliefs about the structure of knowledge. Statements reflecting beliefs about the

Helenrose Fives and Michelle M. Buehl 497

structure of knowledge emerged when individuals discussed the stability of knowledge more so than any other topic in our question set.

Specific topics – what changes? The specificity of knowledge change was another trend that we noted in preservice and practicing teachers’ articulation of beliefs about the stability of teaching knowledge. Some participants indicated that changes in knowledge would vary depending on the aspect of knowledge under consideration. These specific views of knowledge change as well as the diversity of responses are reflected in statements from a practicing teacher:

Content area may change quite a bit – classes offered will change some – but the core skills to be good teacher – passion, ability to work and collaborate with students and peers, classroom management – will not change much. (ID 423)

Participants indicated a variety of topics that will or will not experi­ence change in the next twenty years. Topics indicated included: content or subject area knowledge, human development and learning theory, meeting student needs, specific teaching methods (e.g., classroom man­agement), general teaching methods, technology in education, issues in education, and communication skills. These topics are reflective of the aspects of knowledge needed for teaching discussed in the first frame presented in this chapter. Participants differed in their expectations of which topics would change and how (e.g., amount, direction, and qual­ity) those changes would manifest.

Reason for change. Participants sometimes indicated a rationale for the change or for the lack of change anticipated. We identified two overarching explanations for change (or lack thereof): topic as source of change and environment as source of change. Technology, content, teaching, development, and learning theories were offered as topics that would lead to change or not. For instance, it was indicated that as new discoveries are made knowledge will change. Alternatively, it was also suggested by some that the knowledge core was already established and no change would occur. Participants also described changes in school contexts, families, policy, and the overarching culture as reasons for a changing or stable body of knowledge related to teaching.

Implications for beliefs about the stability of teaching knowledge. Our data provided various views about the stability of teaching knowledge. Whereas some indicated teaching knowledge is static and will not change (e.g., a more absolutist stance; Kuhn, 1991), others suggested that teach­ing knowledge will change to varying degrees. For instance, when the preservice teacher quoted earlier (i.e., ID 108) stated that “I know [teach­ing knowledge] will change, everything does” she acknowledges the “legitimate uncertainty in the world” (Moore, 2002, p. 20), or at least in

Teachers’ articulation of beliefs about teaching498

relation to teaching knowledge, indicative of a more multiplistic stance (e.g., Kuhn, 1991; Perry, 1970). Such varied views have implications for teacher education and development and may account for some of the attitudes and behaviors of preservice and practicing teachers.

For instance, individuals who believe teaching knowledge will change may recognize the importance of staying up­to­date on the current research and participating in development opportunities. Alternatively, such individuals could adopt the stance that since the knowledge will ultimately change, it is not worth learning. Similarly, those who believe knowledge will not change may devalue and avoid opportunities to learn new information or they may seek out information in an area they are new to for the purpose of learning the “truth.”

As discussed in developmental models of epistemic cognition (e.g., King and Kitchener, 1994; Kuhn, 1991; Perry, 1970), beliefs about the certainty of knowledge and knowing are related to their views of how to evaluate the “correctness” or appropriateness of alternative views and changing knowledge. Consequently, an emphasis on the uncertainty of the knowledge needed for teaching may not be beneficial unless pre­service and practicing teachers are also encouraged to view themselves as active participants in the construction of meaning based on critical evaluation of available evidence and theory (i.e., take a more evalutiv­ist stance; e.g., King and Kitchener, 1994; Kuhn, 1991; Perry, 1970). Teacher educators should provide preservice and practicing teachers with experience in how alternative and “new” information can be eval­uated in light of existing theory and research.

Beliefs about the stability of knowledge may indirectly affect classroom practice through teachers’ awareness of new methods and approaches to teaching. Those who do not seek out new knowledge about teaching certainly cannot employ it in the classroom. Additionally, beliefs about the stability of knowledge within specific content areas (e.g., science) may also place students at a disadvantage if teachers are not aware of recent developments in the field or if teachers encourage students to adopt more static views of knowledge within particular fields. Thus, it may be beneficial for teacher educators to highlight how some aspects of teaching knowledge are changing more rapidly (e.g., content knowl­edge), whereas others change more gradually (e.g., understanding of learning and development) or in a more iterative fashion (e.g., “new” instructional practices; Alexander et al., 1996).

Our data suggest that preservice and practicing teachers already view different aspects of teaching knowledge to be more or less stable. This finding further underscores the need to treat teaching as a domain of study and to assess beliefs about the stability of knowledge relative to

Helenrose Fives and Michelle M. Buehl 499

specific aspects of teaching knowledge. Such a fine­grained analysis may be needed to fully explore and understand the implications of these beliefs for teacher development and practice.

Domain qualities and skills: what qualities or skills do teachers need to have?

It is a teacher’s duty to provide support, educational and emotional, to students so they can learn to the best of their ability. (ID 421, practicing teacher)

Although our questions focused on the knowledge and ability needed for teaching, across the responses participants repeatedly reported skills or qualities teachers needed that we saw as distinct from knowledge. We organized these skills and qualities into five emergent themes: com-munication, adaptability and ingenuity, nurturance and care, enthusiasm, and integrity and commitment. Interestingly, as we will discuss, the skills and qualities revealed by our participants reflect attitudes, beliefs, and behaviors akin to the dispositions referred to by the National Council for the Accreditation of Teacher Education (NCATE).

Communication. Among the responses there was an overwhelming articulation of the need for teachers to have communication skills. Specifically, the ability to communicate with all members of the educa­tional community (e.g., parents, students, colleagues, supervisors, etc.) was remarked upon repeatedly. In addition, participants referred to the ability to orate and explain as well as the ability to establish rapport, build relationships, and listen.

Adaptability and ingenuity. The need for ingenuity and adaptabil­ity in teachers emerged as a second relevant theme within this frame. Throughout the responses participants indicated that teachers needed to be creative, flexible, organized, comfortable in schools, and able to think on the spot and adapt.

Nurturance and care. The remaining three themes emerged related to more affective qualities. The first of these more affective themes pertained to nurturance and care. Participants repeatedly reported that teachers need to care, be kindhearted, demonstrate compassion, empa­thy, patience, sympathy, and respect for each student. There was a focus in these responses on the conceptualization of teaching as more than relaying information and a recognition that not everyone is meant to be a teacher. This perspective was articulated by one preservice teacher:

I feel that not everyone has the heart to be a teacher. And having the heart makes the teacher. I am sure anyone can stand up and instruct, but having the patience, charisma, and ability to reach children is not easy. (ID 108)

Teachers’ articulation of beliefs about teaching500

Enthusiasm. The second affective quality that emerged from our par­ticipants’ responses was a need for teachers to demonstrate enthusiasm for their job, for learning, and for their content matter. The underlying mes­sage in these statements seemed to be that unless teachers are motivated and enthusiastic about their work, they cannot motivate their students.

Integrity and commitment. The final set of affective qualities empha­sized integrity, the need for teachers to be fair, and commitment. Integ­rity was articulated by one practicing teacher: “One must understand that fair does not mean that everyone is treated the same. Fair means that each child gets what he/she needs” (ID 141). Similarly, there was an emphasis on the need for teachers to want to make a difference and be committed to teaching. In the following quote a preservice teacher articulates the integrated need for commitment, and integrity to the teaching profession and highlights the perspective that not everyone should be a teacher:

Not everyone is cut out to be a teacher. [Teaching] takes a huge amount of time and commitment. Teaching is a chance for you to change the lives of so many children you come in contact with each year. Kids tend to model what they see – so teachers must be very careful with what they say and how they dress. (ID 114)

Implications of qualities and skills. We were initially surprised by the emergence of this frame and its related themes. That is, our questions were focused on the knowledge needed for teaching and beliefs about the ability to teach. However, the fact that skills and affective qualities were reported by our participants provides evidence of the complexity of preservice and practicing teachers’ belief systems and the interrela­tions among knowledge and beliefs (e.g., Pajares, 1992). Although some have noted that reference to the affective aspects of teaching are more prevalent among preservice and novice teachers (e.g., Hollingsworth, 1989; Weinstein, 1998), we noted responses related to this frame across participants. Additional research is needed to see if certain skills or affective qualities are reported more or less by individuals of varying levels of teaching experience.

As previously noted, the skills and qualities identified by our partici­pants seem reflective of what the NCATE refers to as dispositions. That is, the first NCATE standard requires that teacher candidates “dem­onstrate the content, pedagogical, and professional knowledge, skills, and dispositions necessary to help all students learn” (NCATE, 2006, p. 10). The glossary of the document defines dispositions as:

The values, commitments, and professional ethics that influence behaviors toward students, families, colleagues, and communities and affect student

Helenrose Fives and Michelle M. Buehl 501

learning, motivation, and development as well as the educator’s own profes­sional growth. Dispositions are guided by beliefs and attitudes related to values such as caring, fairness, honesty, responsibility, and social justice. (NCATE, 2006, p. 53)

Our data suggest that participants believe such dispositions may be important to success in teaching. However, the prospect of selecting, identifying, and assessing these skills and qualities as part of teacher education curriculum and assessment plans is troubling on several levels. Debates have emerged regarding the beliefs that should be endorsed and the conceptualization of the “right” kind of person to be a teacher (e.g., McKnight, 2004; Damon, 2005). Also, there is the issue of whether dispositions are used to screen out potential teacher candi­dates before entering a teacher education program, or if teacher candi­dates are expected to develop the dispositions throughout the course of the program. Further, even if agreement is met on a set of dispositions and how to use them, how does one assess such dispositions? Defin­ing and assessing personal qualities, such as caring, may lead teacher education programs and certification bodies into murky waters. The skills and qualities articulated by our participants may shed another perspective on the role of dispositions in teacher preparation and help to focus on the dispositions that seem most relevant to current and future teachers.

Conclusions and implications

If beliefs lie at the heart of teachers’ practice, then understanding the relations among teachers’ beliefs, education, and classroom practices is essential. Yet, examining the relations among constructs requires that one can assess those constructs in a meaningful way. As noted within the literature, teacher beliefs are a particularly “messy construct” (Pajares, 1992).

As a case in point, at the beginning of this chapter and at the intro­duction to each of our five frames, we included the response from one practicing female teacher who has been teaching second grade for five years. Her responses provide further evidence of the complexity of beliefs. That is, this teacher simultaneously possesses beliefs about teaching knowledge, the ability to teach, the source, and stability of teaching knowledge, and qualities recognized as important for teach­ers to possess. Such beliefs are likely to be interrelated. For instance, there may be connections between individuals’ beliefs about the ability to teach and their beliefs about the source of teaching knowledge. The

Teachers’ articulation of beliefs about teaching502

practicing teacher we feature mentioned the role of “staff development” when queried about the ability to teach and the source of teaching knowledge. She may also hold that some aspects of teaching knowledge (e.g., classroom management, special populations, or research­based practices) are acquired through staff development whereas others are learned through observing master teachers. These potential relations underscore the perspective that beliefs do not exist in isolation and that they may simultaneously and reciprocally exert influence on each other and the actions of the individual. These multiple connections among the components of individuals’ belief systems offer both potential ave­nues for and impediments to changing teachers’ knowledge and beliefs (e.g., Schommer­Aikins, 2004).

Additionally, the responses offered by this teacher illustrate issues related to the consistency of beliefs as well as individuals’ awareness of their beliefs. This practicing teacher is relatively articulate in expressing her beliefs, there is consistency across her responses, and her responses are reflected in each of the emergent themes. However, she was not probed individually about her responses. She, and other participants, may hold additional beliefs or even contradictory implicit beliefs that were not uncovered by our questions. Thus, it is difficult to determine which beliefs are most prevalent or influential on preservice and prac­ticing teachers’ cognitions and behaviors. Additional research is needed to fully understand the influence of these beliefs.

The framework we presented represents our initial attempts to under­stand the nature of preservice and practicing teachers’ beliefs about teaching knowledge and teaching ability by conceptualizing teach­ing knowledge as a professional domain of knowledge and employ­ing frameworks similar to those used to study students’ beliefs about knowledge and the ability to learn. We examined beliefs with respect to what constituted teaching knowledge, its source and stability, and beliefs about the ability to teach. The overarching goals of this work are threefold: (1) to develop reliable and valid measures for studying teachers’ beliefs; (2) to add to the existing theoretical perspectives on teachers’ beliefs; and (3) to offer some immediate implications for those vested in the teacher education process. Here we expand on our con­cerns and suggestions for future research.

Methodological concerns and suggestions

One of our intentions is to use the insight we have gained through the research described to develop reliable and valid measures for study­ing teachers’ beliefs about teaching knowledge and teaching ability in

Helenrose Fives and Michelle M. Buehl 503

relation to teacher education, motivation, and practice. However, our work thus far has raised several methodological issues related to the assessment of teachers’ beliefs.

Specifically, our analysis of individuals’ response data illustrated the complexity, specificity, and variations in teachers’ beliefs. Across the emergent frames, our respondents articulated various perspectives and beliefs. For instance, there was considerable variation in individuals’ conceptualizations of the domain of teaching knowledge. Although most participants referred to more than one aspect of teaching knowl­edge, underscoring the complexity of teaching as a domain of knowl­edge, no participant referred to all of the themes we identified. The variations in what is considered teaching knowledge certainly provide insight into what preservice and practicing teacher value and view as necessary for teaching. However, such distinctions also have methodo­logical implications. When assessing preservice or practicing teachers’ beliefs about teaching knowledge, one needs to understand how each individual is conceptualizing the domain and weighting the value or importance of each aspect of knowledge within the domain.

This may require a two­pronged methodological attack for quantita­tive research. First, we must constrain the domain sufficiently to be reasonably sure respondents are considering the same body of knowl­edge. Our results suggest that teaching knowledge should be treated as a specific domain of knowledge. Further, our first frame offers areas of additional lines of demarcation within the domain of teaching knowl­edge. Once items specific to the different aspects of teaching knowl­edge are developed researchers can evaluate the interpretation of such items through read­aloud protocols. Second, we must get a measure of how much respondents value each aspect of teaching knowledge. Per­haps, respondents can quantitatively rank order or otherwise indicate the percentage of value ascribed to each body of knowledge within the domain.

Although there were variations across individuals, we also docu­mented evidence of variations within individuals that suggested greater specificity of beliefs as well as individuals’ ability to hold multiple, and perhaps contradictory, perspectives. For instance, individuals endorsed different beliefs about the source and stability of knowledge depending on the aspect of teaching knowledge (e.g., content knowledge, instruc­tional knowledge) under consideration. Thus, when developing meas­ures related to the source or stability of knowledge it may be useful to assess beliefs relative to the specific aspects of teaching knowledge.

As another example of the nuances of individuals’ beliefs, an indi­vidual who believes that the ability to teach is innate for some but

Teachers’ articulation of beliefs about teaching504

learned for others may be classified as holding innate or learned ability beliefs. However, this individual may approach teaching and learn­ing differently from those with solely innate or learned views as well as those who believe in an innate teaching ability that is improved over time. Consequently, how these beliefs are assessed may affect the apparent relations to other beliefs and practices. Assessment techniques need to be sensitive enough to detect such nuances in individuals’ beliefs.

In our work, we have developed a Likert scale measure with items reflecting the various types of ability beliefs we identified in our par­ticipants’ responses (Fives and Buehl, 2005). Factor analysis of data gathered from 351 participants identified four separate factors reflec­tive of the themes we identified in the open­ended responses. Factors emerged representing beliefs in: innate teaching ability, the ability to teach being innate for some and learned for others, the ability to teach as an innate talent that requires training and polish, and a belief that the ability to teach can be learned. Items related to teaching as a calling or gift were assigned to the innate ability belief factor. As noted in our discussion of these beliefs (innate and calling), a theoretical distinction can be made. However, in both instances the source of teaching abil­ity was beyond the control of the individual, be it from genetics or a higher being. This measure may allow researchers to explore how more nuanced beliefs about the ability to teach are related to other teacher beliefs and actions.

We hold that some of the incongruence that has been identified between teachers’ beliefs and practices (e.g., Levitt, 2001; Olafson and Schraw, 2006) may be due to differences in practitioners’ and research­ers’ conceptualizations of the beliefs of interest. For instance, Olafson and Schraw (2006) reported that of the practicing teachers in their study “none of the participants indicated support for the realist position, yet all of their final projects included instructional practices that were coded as realist” (p. 78). These practices were coded by the researchers as realist however it may be that these teachers do not see the identified practices as being realist in nature, thus the distinctions in beliefs may be due more to differences in definitions than beliefs.

Alternatively, we must also recognize that the “beliefs” expressed in interviews or indicated on surveys may not reflect preservice or practic­ing teachers’ “true” perspectives. That is, teacher education programs may not be changing preservice teachers’ beliefs. Instead preservice and practicing teachers may merely adopt the rhetoric conferred by teacher education programs to explain and act on their initially held perspec­tives (Holt­Reynolds, 2000). Alternatively, preservice and practicing

Helenrose Fives and Michelle M. Buehl 505

teachers’ understanding and appropriation of terms (e.g., construc­tivism, active learning) may differ from the understandings held by researchers and teacher educators. Furthermore, researchers need to be sensitive to the role that organizational constraints may play in teach­ers’ implementation of their beliefs (e.g., Schraw and Olafson, 2002; Wilcox­Herzog, 2002).

Another issue to consider in the examination of the relations between beliefs and practice is that at any point the preservice or practicing teachers’ beliefs may be unexamined, under development, or in transition. Indeed, previous research has noted discernable differences in beliefs across individuals who are entering a teacher education program, com­pleting a field experience, entering the classroom as a novice teacher, and persisting in the teaching profession (Patrick and Pintrich, 2001; Richardson, 1996). Within the personal epistemology, there is evidence of a developmental trajectory that individuals’ beliefs may proceed upon (e.g., King and Kitchener, 1994; Kuhn, 1991; Perry, 1970). Thus, preservice and practicing teachers may not have multiple incongruent beliefs, rather it may be that they are still working through their beliefs on a particular topic and fitting that perspective into their larger belief framework.

Future research needs to examine the developmental trajectory of teaching beliefs as a domain. It may be, as with Alexander’s (1997) model of domain learning, which identifies unique knowledge, strategy use, and motivation at varying points in one’s academic development, that there are also distinct developmental domain­specific epistemic beliefs at play as well. Identifying and understanding the beliefs that may be more or less adaptive at different points in a teachers’ profes­sional development is an empirical question that begs future research. One could argue that beliefs about teaching knowledge as stable and certain may help preservice and new teachers to frame their professional domain and establish as sense of competence that will facilitate their early teaching practices. Such beliefs may also be adaptive for practic­ing teachers who are balancing work, family, and personal interests. In contrast, individuals may perceive uncertain and changing knowledge as an opportunity for continued growth and the possibility of identify­ing alternatives to teaching practices they find ineffective.

Further, it may be beneficial to examine the correspondence between individuals’ developmental level with respect to beliefs about knowl­edge, learning, and teaching and their actual level of experience and practices in the classroom. For instance, Muis et al. (2006) raised the question of how teachers with naïve views about knowledge can foster more sophisticated views in their students.

Teachers’ articulation of beliefs about teaching506

Theoretical perspective

The research presented here offers an initial attempt to bridge the theo­retical fields of teachers’ beliefs and students’ epistemic belief systems through a data­based approach that gives voice to the perspectives of those in the intended participant and intervention pool. The emergent themes we presented (i.e., domain knowledge, source of knowledge, ability beliefs, stability of knowledge, and domain qualities and skills; Table 15.1) reflect existing theoretical frameworks about epistemic beliefs (e.g., Buehl and Alexander, 2005; Hofer, 2000; Schraw et al., 2002), teaching knowledge (e.g., Elbaz, 1983; Grossman, 1990; Rich­ardson, 1996; Shulman, 1987), and ability beliefs (e.g., Dweck, 2002; Schommer, 1990).

Our framework reveals that preservice and practicing teachers hold perspectives about teaching knowledge that have been seen in other populations (e.g., students) regarding academic domains of knowl­edge. For instance, our framework reflects beliefs about the source and stability of teaching knowledge that may vary depending on the aspect of teaching knowledge under consideration. In particular, our study highlights that preservice and practicing attribute their teach­ing knowledge to the multiple sources. Similarly, our participants held varied perspectives about the stability of knowledge within the domain of teaching knowledge itself. For instance, some participants reported that some aspects of knowledge will change (e.g., teaching methods) and others will not (e.g., developmental theories).

Our framework also addresses beliefs about the ability to teach and how that ability is derived. Akin to previous work on ability beliefs (e.g., Dweck, 2002), our framework recognizes the distinction between fixed or innate ability beliefs and learned ability beliefs. However, our frame­work adds additional theoretical dimensions by elaborating on the vari­ations in beliefs about the ability to teach that might exist. Further, the interrelations among beliefs about the origins of one’s teaching abil­ity and the sources of teaching knowledge reflect the interconnections among beliefs about knowledge and learning as suggested by Schommer­ Aikins (2004).

Researchers of epistemic beliefs must begin to account for these more complex perspectives within their theories. Treating teaching knowledge as a domain of knowledge with subcomponents for which individuals may possess different beliefs would be one way to take the complexity into account. The present work also illuminates the per­spective that beliefs about ability may also be domain or even task spe­cific in nature and that the various aspects of individuals belief systems

Helenrose Fives and Michelle M. Buehl 507

need to be considered in relation to one another. Such calls for the examination of beliefs at a domain­specific level as well as to examine the interrelations among beliefs within and across domains have been made by others as well (e.g., Buehl and Alexander, 2001; Muis et al., 2006; Schommer­Aikins, 2004).

Recommendations and implications for teacher education

The framework presented here allows teacher educators and those fos­tering professional development of teachers an opportunity to pinpoint potential motivators and demotivators in preservice and practicing teachers’ belief systems. For example, the variations in teaching knowl­edge identified as needed for teaching may suggest that learners may be more or less willing to learn about specific topics based on the perceived utility value of that knowledge (Wigfield and Eccles, 2000). Our work also highlights the sources of knowledge valued and accepted by pre­service and practicing teachers.

Such findings are useful in the context of promoting conceptual change in preservice and practicing teachers. That is, Patrick and Pintrich (2001) noted the need for instructors to be aware of preservice teachers’ beliefs as well as to provide opportunities for preservice teach­ers to become aware of their beliefs and to explicitly discuss the epis­temic beliefs that may underlie other implicit theories and beliefs. By understanding preservice and practicing teachers’ valuing of the sources of knowledge, teacher educators can present evidence that the preservice and practicing teachers perceive as credible, an important condition for conceptual change (e.g., Chinn and Bewer, 1993; Pintrich et al., 1993).

Additionally, teacher educators could engage students in explicit discussions of the source of teaching knowledge (i.e., how the vari­ous experience in their teacher preparation programs contribute to the development of their knowledge of how to teach). For instance, preservice and practicing teachers often view field experiences as the most important part of the teacher education program (e.g., McIntrye et al., 1996). However, the effectiveness of such experiences is variable depending on factors such as the candidate’s knowledge as well as the nature of the placement and cooperating teacher (e.g., Hamman et al., 2007). Teacher educators may wish to discuss this with preservice teachers and explicate how different assignments and learning experi­ences will contribute to their understanding and effectiveness in the classroom as teachers.

Further, in discussing the uncertainty and changing nature of knowl­edge, teacher educators may also discuss how current understandings

Teachers’ articulation of beliefs about teaching508

came to be accepted as well as underscore the need for preservice and practicing teachers to making meaning and evaluate new ideas and innovations in light of their personal experience and their grounding in the formal knowledge within the field. That is, as teacher educators interact with preservice and practicing teachers, it may be beneficial for the teacher educators to foster an evaluativist stance (e.g., King and Kitchener, 1994; Kuhn, 1991) toward teaching knowledge. Such an emphasis may aid to promote the epistemic development of preservice and practicing teachers as well as contribute to a sense of professional judgment.

In our participants’ responses, there was, to us, a surprising emphasis on the aspects of teaching that are not or cannot be taught. For instance, it is not surprising that a classroom teacher needs to be organized, able to multi­task, and able to manage his/her time effectively. However, few of these necessary skills are ever explicitly taught in teacher prepara­tion programs. The strategies for mutli­tasking, time management, and organization of paperwork are seldom mentioned in the university set­ting. As with other strategies (e.g., Alexander et al., 1998), there is the assumption that learners will “pick it up” rather than recognition that these skills and strategies can and should be explicitly taught. Given the emphasis on these issues in our data these may be areas that need more explicit instruction.

Participants repeatedly remarked on more affective qualities that are needed to be a teacher. Teachers need to love children, be patient, “have heart.” These myriad of perceived qualities needed for effective teaching and their connection to the NCATE standards requiring the assessment of dispositions leads to many methodological and ideologi­cal questions for teacher educators such as: What qualities are needed for teachers to be effective? How are those qualities defined, demon­strated, and assessed? For example, what does it mean for a teacher to care? Can and should these qualities be standardized? What about the need for learners – future adults – to be exposed to many kinds of people, who exhibit a variety of dispositions and attitudes about teach­ing and learning?

Finally, although we asked questions about teaching knowledge and teaching ability we received responses referring to skills, qualities, policy, accreditation, certification, course requirements, rationaliza­tions, and ideological rants, suggesting that these issues are not clearly demarcated in the minds of preservice and practicing teachers. As dis­cussed by others, perhaps one goal of teacher preparation and profes­sional development should be to help teachers systematically examine their beliefs about teaching as they move their beliefs from tacit to

Helenrose Fives and Michelle M. Buehl 509

explicit and from transitional to well­developed and enacted (Gill et al., 2004; Olafson and Schraw, 2006; Patrick and Pintrich, 2001; Richard­son, 1996; White, 2000). However, to do so, researchers and teacher educators need to be sensitive to the nuances in preservice and practic­ing teachers’ beliefs. Additionally, we need a greater understanding of the types of beliefs that are most adaptive to preservice and practic­ing teachers throughout their developmental trajectory as learners and educators.

R EF ER ENCES

Alexander, P. A. (1997). Mapping the multidimensional nature of domain learning: The interplay of cognitive, motivational and strategic forces. In P. R. Pintrich and M. L. Maehr (Eds.), Advances in motivation and achieve-ment (Vol. 10, 213–50). Greenwich, CT: JAI Press.

(1992). Domain knowledge: Emerging themes and emerging concerns. Edu-cational Psychologist, 27, 33–51.

Alexander, P. A., Graham, S., and Harris, K. R. (1998). A perspective on strat­egy research: Progress and prospects. Educational Psychology Review, 10, 131–54.

Alexander, P. A., Murphy, P. K., and Woods, B. (1996). Of squalls and fath­oms: Navigating the seas of educational innovation. Educational Researcher, 25(3), 31–6.

Bendixen, L. D., Schraw G., and Dunkle, M. E. (1998). Epistemic beliefs and moral reasoning. The Journal of Psychology, 132(2), 187–200.

Bogden, R. C. and Biklen, S. K. (1992). Qualitative research for education. Bos­ton: Allyn and Bacon.

Book, C., Byers, J., and Freeman, D. (1983). Student expectations and teacher education traditions with which we can and cannot live. Journal of Teacher Education, 34, 9–13.

Borko, H., Davinroy, K. H., Bliem, C. L., and Cumbo, K. (2000). Exploring and supporting teacher change: Two third­grade teachers’ experiences in a mathematics and literacy staff development project. Elementary School Journal, 100, 273–306.

Brownlee, J. (2004). Teacher education students’ epistemological beliefs: Devel­oping a relational model of teaching. Research in Education, 72, 1–17.

Brownlee, J. and Berthelsen, D. (2006). Personal epistemology and relational pedagogy in early childhood teacher education programs. Early Years, 1, 17–29.

Brownlee, J., Purdie, N., and Boulton­Lewis, G. (2001). Changing epistemo­logical beliefs in pre­service teacher education. Teaching in Higher Educa-tion, 6(2), 247–68.

Buchanan, C. M., Eccles, J. S., Flanagan, C., Midgley, C., Feldlaufer, H., and Harold, R. (1990). Parents’ and teachers’ beliefs about adolescents: Effects of sex and experience. Journal of Youth and Adolescence, 19, 363–94.

Buehl, M. M. and Alexander, P. A. (2001). Beliefs about academic knowledge. Educational Psychological Review, 13, 385–418 (special issue).

Teachers’ articulation of beliefs about teaching510

(2005). Motivation and performance differences among domain­specific epistemological belief clusters. American Educational Research Journal, 42, 697–726.

(2006). Examining the dual nature of epistemological beliefs. International Journal of Educational Research, 45, 28–42.

Buehl, M. M. and Fives, H. (2009). Exploring teachers’ beliefs about teach­ing knowledge: Where does it come from? Does it change? The Journal of Experimental Education, 77(4), 367–408.

Buehl, M. M., Alexander, P. A., and Murphy, P. K. (2002). Beliefs about schooled knowledge: Domain general or domain specific? Contemporary Educational Psychology, 27, 415–49.

Butler, R. (2000). Making judgments about ability: The role of implicit theories of ability in moderating inferences from temporal and social comparison information. Journal of Personality and Social Psychology, 78, 965–78.

Calderhead, J. (1996). Teachers: Beliefs and knowledge. In D. C. Berliner and R. C. Calfee (Eds.), Handbook of educational psychology (709–25). New York: Macmillan.

Carter, K. (1990). Teachers’ knowledge and learning to teach. In W. Houston (Ed.), Handbook of research on teacher education (291–310). New York: Macmillian.

Chinn, C. A. and Brewer, W. F. (1993). The role of anomalous data in knowl­edge acquisition: A theoretical framework and implications for science instruction. Review of Educational Research, 63, 1–49.

Damon, W. (2005). Personality test: The dispositional dispute in teacher prepara-tion today, and what to do about it. Washington DC: The Thomas B. Ford­ham Foundation. Retrieved March 4, 2007, from www.edexcellence.net/detail/news.cfm?news_id=343.

de Corte, E., Op ’t Eynde, P., and Verschaffel, L. (2002). “Knowing what to believe”: The relevance of students’ mathematical beliefs for mathemat­ics education. In B. K. Hofer and P. R. Pintrich (Eds.), Personal episte-mology: The psychology of beliefs about knowledge and knowing (297–320). Mahwah, NJ: Lawrence Erlbaum Associates.

diSessa, A. A. (1993). Toward an epistemology of physics. Cognition and Instruc-tion, 10, 105–225.

Dweck, C. (2002). The development of ability conceptions. In A. Wigfield and J. Eccles (Eds.), The development of achievement motivation. San Diego, CA: Academic Press.

Dweck, C. S. and Molden, D. C. (2005). Self­theories: Their impact on com­petence motivation and acquisition. In A. J. Elliot and C. S. Dweck (Eds.), Handbook of competence and motivation (122–40). New York: Guilford Press.

Dweck, C. S., Chiu, C., and Hong, Y. (1995). Implicit theories and their role in judgments and reactions: A world from two perspectives. Psychological Inquiry, 6, 267–85.

Eccles, J. and Wigfield, A. (2002). Motivational beliefs, values, and goals. Annual Review of Psychology, 53(1), 109–32.

Elbaz, F. (1983). Teacher thinking: A study of practical knowledge. New York: Nichols.

Helenrose Fives and Michelle M. Buehl 511

Fives, H. and Buehl, M. M. (2005). Exploring teachers’ pedagogical beliefs: Devel-oping a measure. Invited paper presented at the regional paper award session of the annual meeting of the American Educational Research Association, Montreal, Canada (April).

(2008). What do teachers believe? Developing a framework for examining beliefs about teachers’ knowledge and ability. Contemporary Educational Psychology, 33(2), 134–76.

Flores, B. B. (2001). Bilingual education teachers’ beliefs and their relation to self­reported practice. Bilingual Research Journal, 25, 275–99.

Flowerday, T. and Schraw, G. (2000). Teacher beliefs about instructional choice: A phenomenological study. Journal of Educational Psychology, 92, 634–45.

Gill, M. G., Ashton, P. T., and Algina, J. (2004). Changing preservice teachers’ epistemological beliefs about teaching and learning in mathematics: An intervention study. Contemporary Educational Psychology, 29, 164–85.

Glaser, B. G. and Strauss, A. L. (1967). The discovery of grounded theory: Strate-gies for qualitative research. Chicago: Aldine.

Grossman, P. L. (1990). The making of a teacher: Teacher knowledge and teacher education. New York: Teachers College Press.

Hamman, D., Fives, H., and Olivarez, A. Jr. (2007). Efficacy and pedagogical interaction in cooperating and student teacher dyads. Journal of Classroom Interaction, 41/42, 55–63.

Hammer, D. (1994). Epistemological beliefs in introductory physics. Cognition and Instruction, 12, 151–83.

Henderson, V. and Dweck, C. S. (1990). Motivation and achievement. In S. S. Feldman and G. R. Elliott (Eds.), At the threshold: The developing adolescent (308–29). Cambridge, MA: Harvard University Press.

Hofer, B. K. (2000). Dimensionality and disciplinary differences in personal epistemology. Contemporary Educational Psychology, 25, 378–405.

(2001). Personal epistemology research: Implications for learning and teach­ing. Educational Review, 13, 353–83.

(2002). Personal epistemology as a psychological and educational con­struct: An introduction. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (3–14). Mahwah, NJ: Lawrence Erlbaum Associates.

(2004). Exploring the dimensions of personal epistemology in differing classroom contexts: Student interpretations during the first year of col­lege. Contemporary Educational Psychology, 29, 129–63.

(2006). Domain­specificity of personal epistemology: Resolved questions, persistent issues, new models. International Journal of Educational Research, 45, 57–70.

Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learn­ing. Review of Educational Research, 67, 88–140.

Hollingsworth, S. (1989). Prior beliefs and cognitive change in learning to teach. American Educational Research Journal, 26, 160–89.

Holt­Reynolds, D. (2000). What does the teacher do? Constructivist pedago­gies and prospective teachers’ beliefs about the role of a teacher. Teaching and Teacher Education, 16, 21–32.

Teachers’ articulation of beliefs about teaching512

Hong, Y. and Dweck, C. S. (1992). Implicit theories as predictors of self-inference processes. Paper presented at the annual convention of the American Psy­chological Society, San Diego (June).

Jehng, J. J., Johnson, S. D., and Anderson, R. C. (1993). Schooling and stu­dents’ epistemological beliefs about learning. Contemporary Educational Psychology, 18, 23–35.

Johnson, R. J. (2000). Case studies of expectation climate at two bilingual edu­cation schools. Bilingual Research Journal, 24, 261–76.

Jordan, A. and Stanovich, P. (2003). Teachers’ personal epistemological beliefs about students with disabilities as indicators of effective teaching practice. Journal of Research in Special Educational Needs, 3, 1–14.

Jordon, A., Lindsay, L., and Stanovich, P. (1997). Classroom teachers’ instruc­tional interactions with students who are exceptional, at risk, and typically achieving. Remedial and Special Education, 18(2), 82–93.

Kagan, D. M. (1992). Implications of research on teacher belief. Educational Psychologist, 27, 65–90.

Kemple, K. M. (1996). Teachers’ beliefs and reported practices concern­ing sociodramatic play. Journal of Early Childhood Teacher Education, 17, 19–31.

King, P. M. and Kitchener, K. S. (1994). Developing reflective judgment: Under-standing and promoting intellectual growth and critical thinking in adolescents and adults. San Francisco: Jossey­Bass.

(2002). The reflective judgment model: Twenty years of research on epis­temic cognition. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing (37–62). Mahwah, NJ: Lawrence Erlbaum Associates.

Kowalski, K., Pretti­Frontczak, K., and Johnson, L. (2001). Preschool teach­ers’ beliefs concerning the importance of various developmental skills and abilities. Journal of Research in Childhood Education, 16, 5–14.

Kuhn, D. (1991). The skill of argument. Cambridge: Cambridge University Press.

Kuhn, D. and Weinstock, M. (2002). What is epistemological thinking and why does it matter? In B. K. Hofer and P. R. Pintrich (Eds.), Personal Epis-temology: The psychology of beliefs about knowledge and knowing (121–44). Mahwah, NJ: Lawrence Erlbaum Associates.

Lawless, K. A. and Smith, E. V. (1997). Teacher beliefs about instructional media: Confirmatory factor structure. International Journal of Instructional Media, 24, 191–6.

Lee, K. (1996). A study of teacher responses based on their conceptions of intelligence. Journal of Classroom Interaction, 31, 1–12.

Leinhardt, G. and Smith, D. (1985). Expertise in mathematics instruc­tion: Subject matter knowledge. Journal of Educational Psychology, 77, 241–71.

Levitt, K. E. (2001). An analysis of elementary teachers’ beliefs regarding the teaching and learning of science. Science Education, 86, 1–22.

Mantzicopoulos, P. Y. and Neuharth­Pritchett, S. (1998). Transitional first­grade referrals: An analysis of school­related factors and children’s char­acteristics. Journal of Educational Psychology, 90, 122–33.

Helenrose Fives and Michelle M. Buehl 513

McAllister, G. and Irvine, J. J. (2002). The role of empathy in teaching cultur­ally diverse students: A qualitative study of teachers’ beliefs. Journal of Teacher Education, 53, 433–43.

McIntrye, D. J., Byrd, D. M., and Foxx, S. M. (1996). Field and laboratory experiences. In J. Sikula (Ed.), Handbook of research on teacher education (171–193). New York: Macmillian.

McKnight, D. (2004). An inquiry of NCATE’s move into virtue ethics by way of dispositions (Is this what Aristotle meant?). Educational Studies, 35, 212–30.

McLachlan­Smith, C. J. and St­George, A. M. (2000). “Children learn by doing”: Teachers’ beliefs about learning, teaching and literacy in New Zealand kindergartens. New Zealand Journal of Educational Studies, 35, 37–47.

Muis, K. R., Bendixen, L. D., and Haerle, F. C. (2006). Domain­generality and domain­specificity in personal epistemology research: Philosophical and empirical reflections in the development of a theoretical framework. Educational Psychology Review, 18(3), 3–54.

Munby, H., Russell, T., and Martin, A. (2001). Teachers’ knowledge and how it develops. In V. Richardson (Ed.), Handbook of research on teaching (4th edn.), 877–904). Washington, DC: AERA.

NCATE (2006). Professional standards for the accreditation of schools, colleges, and departments of education (2006 Edition). National Councils for Accredita­tion of Teacher Education. Retrieved March 4, 2007, from www.ncate.org/ public/standards.asp?ch=4.

Olafson, L. and Schraw, G. (2006). Teachers’ beliefs and practices within and across domains. International Journal of Educational Research, 45(1–2), 71–84.

Op ’t Eynde, P., de Corte, E., and Verschaffel, L. (2006). Epistemic dimen­sions of students’ mathematics­related belief systems. International Journal of Educational Research, 45, 57–70.

Oser, F. K. and Baeriswyl, F. J.(2001). Choreographies of teaching: Bridging instruction to learning. In V. Richardson (ed.), Handbook of Research on Teaching (4th edn., 1031–65). Washington, DC: AERA.

Pajares, M. F. (1992). Teachers’ beliefs and educational research: Cleaning up a messy construct. Review of Educational Research, 62, 307–32.

Paris, S. G., Lipson, M. Y., and Wixson, K. K. (1983). Becoming a strategic reader. Contemporary Educational Psychology, 8, 293–316.

Patrick, H. and Pintrich, P. R. (2001). Conceptual change in teachers’ intuitive conceptions of learning, motivation, and instruction: The role of motiva­tional and epistemological beliefs. In B. Torff and R. J. Sternberg (Eds.), Understanding and teaching the intuitive mind: Student and teacher learning (117–43). Mahwah, NJ: Lawrence Erlbaum Associates.

Perry, W. G. (1970). Forms of intellectual and ethical development in the college years: A scheme. New York: Holt, Rinehart and Winston.

Pintrich, P. R., Marx, R. W., and Boyle, R. A. (1993). Beyond cold conceptual change: The role of motivational beliefs and classroom contextual factors in the process of conceptual change. Review of Educational Research, 63(2), 167–99.

Teachers’ articulation of beliefs about teaching514

Pintrich, P. R. and Schrauben, B. (1992). Students’ motivational beliefs and their cognitive engagement in the classroom academic tasks. In D. Schunk and J. Meece (Eds.), Students perceptions in the classroom (149–83). Hills­dale, NJ: Lawrence Erlbaum Associates.

Pohan, C. A. (1996). Preservice teachers’ beliefs about diversity: Uncovering factors leading to multicultural responsiveness. Equity and Excellence in Education, 29, 62–9.

Qian, G. and Alverman, D. (1995). Role of epistemological beliefs and learned helplessness in secondary school students’ learning science concepts from text. Journal of Educational Psychology, 87, 282–92.

Ravindran, B., Greene, B. A., and Debacker, T. K. (2005). Predicting preserv­ice teachers’ cognitive engagement with goals and epistemological beliefs. Journal of Educational Research, 98(4), 222–32.

Richardson, V. (1996). The role of attitudes and beliefs in learning to teach. In J. Sikula (Ed.), Handbook of research on teacher education (102–19). New York: Macmillian.

Richardson, V. and Placier, P. (2001). Teacher change. In V. Richardson (Ed.), Handbook of research on teaching (905–47). Washington, DC: American Educational Research Association.

Roderick, M. (1994). Grade retention and school dropout: Investigating the association. American Educational Research Journal, 31, 729–59.

Sahin, C., Bullock, K., and Stables, A. (2002). Teachers’ beliefs and practices in relation to their beliefs about questioning at Key Stage 2. Educational Studies, 28, 371–84.

Schoenfeld, A. (1992). Learning to think mathematically: Problem­solving, metacognition, and sense making in mathematics. In D. A. Grouws (Ed.), Handbook of research on mathematics teaching and learning (334–70). New York: Macmillan.

Schommer, M. (1990). Effects of beliefs about the nature of knowledge on comprehension. Journal of Educational Psychology, 82, 498–504.

Schommer­Aikins, M. (2004). Explaining the epistemological belief sys­tem: Introducing the embedded systematic model and coordinated research approach. Educational Psychologist, 39, 19–29.

Schraw, G. and Olafson, L. (2002). Teachers’ epistemological world views and educational practices. Issues in education: Contributions from Educational Psychology, 8(2), 99–148.

(2006). Teachers’ beliefs and practices within and across domains. Interna-tional Journal of Educational Research, 45, 71–84.

Schraw, G., Bendixen, L. D., and Dunkle, M. E. (2002). Development and val­idation of the Epistemic Belief Inventory (EBI). In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowl-edge and knowing (261–75). Mahwah, NJ: Lawrence Erlbaum Associates.

Shulman, L. S. (1987). Knowledge and teaching: Foundations of the new reform. Harvard Educational Review, 57, 1–22.

Sinatra, G. M. and Kardash, C. M. (2004). Preservice teachers’ epistemologi­cal beliefs, dispositions, and views on teaching as persuasion. Contempo-rary Educational Psychology, 29, 483–8.

Helenrose Fives and Michelle M. Buehl 515

Stipek, D. and MacIver, D. (1989). Developmental change in children’s assess­ment of intellectual competence. Child Development, 60, 521–38.

Stipek, D. J., Givvin, K. B., Salmon, J. M., and MacGyvers, V. L. (2001). Teachers’ beliefs and practices related to mathematics instruction. Teach-ing and Teacher Education, 17, 213–26.

Strauss, A. and Corbin, J. (1998). Basics of qualitative research: Techniques and procedures for developing grounded theory (2nd edn.). Thousand Oaks, CA: Sage Publications, Inc.

Tsai, C. C. (2002). Nested epistemologies: Science teachers’ beliefs of teach­ing, learning and science. International Journal of Science Education, 24, 771–83.

(2007). Teachers’ scientific epistemological views: The coherence with instruction and students’ views. Science Education, 91, 222–43.

Tschannen­Moran, M. and Woolfolk­Hoy, A. (2001). Teacher efficacy: Cap­turing an elusive construct. Teaching and Teacher Education, 17, 783–805.

Weiner, B. (1985). An attributional theory of achievement and motivation. Psy-chological Review, 92, 548–73.

Weinstein, C. S. (1998). “I want to be nice, but I have to be mean”: Explor­ing prospective teachers’ conceptions of caring and order. Teaching and Teacher Education, 14, 153–63.

White, B. C. (2000). Pre­service teachers’ epistemology viewed through per­spectives on problematic classroom situations. Journal of Education for Teaching, 26, 279–305.

Wigfield, A. and Eccles, J. S. (2000). Expectancy­value theory of achievement motivation. Contemporary Educational Psychology, 25, 68–81.

Wigfield, A., Eccles, J. S., Schiefele, U., Roeser, R., and Davis­Kean, P. (2006). Development of achievement motivation. In N. Eisenberg et al. (Eds.), Handbook of child psychology, Vol. 3: Social, emotional, and personality devel-opment (6th edn., 933–1002). Hoboken, NJ: John Wiley and Sons, Inc.

Wilcox­Herzog, A. (2002). Is there a link between teachers’ beliefs and behav­iors? Early Education and Development, 13(1), 81–106.

Woolfolk­Hoy, A., Davis, H., and Pape, S. J. (2006). Teacher knowledge and beliefs. In P. A. Alexander and P. H. Winne (Eds.), Handbook of educa-tional psychology (2nd edn., 715–37). Mahwah, NJ: Lawrence Erlbaum Associates.

516

16 Beyond epistemology: assessing teachers’ epistemological and ontological worldviews

Lori Olafson and Gregory Schraw

University of Nevada, Las Vegas

An important issue in epistemological research concerns the meas­urement of epistemological beliefs and other related methodological issues (Hofer, 2002; Pintrich, 2002). In our work, we have been inter­ested in measuring the epistemological beliefs of practicing teachers and describing the relationship between epistemological beliefs and instructional practices. Previously, we documented lack of alignment between teachers’ epistemological world views and their teaching practices (Schraw and Olafson, 2002; Olafson and Schraw, 2002). In these studies we found, for example, that the majority of our forty­two participants endorsed student­centered instructional practices and believed that learners must construct shared understandings in sup­portive contexts in which teachers serve as facilitators. Yet these teach­ers also reported using teacher­centered instructional practices such as whole­group completion of common work sheets. These findings of poor alignment between teachers’ beliefs and practices are consistent with previous studies (Levitt, 2001; Marra, 2005; White, 2000).

Two issues are of concern regarding research examining alignment between teachers’ beliefs and practices. The first issue is a concep­tual shortcoming due to focusing on epistemology without regard to ontology. Unlike epistemological beliefs, few studies have examined teachers’ ontological beliefs, nor have any studies investigated the joint contribution of epistemological and ontological beliefs. It is our belief that teachers’ epistemological (i.e., beliefs about the nature and acquisi­tion of knowledge) beliefs must be examined in conjunction with their ontological beliefs (i.e., beliefs about the nature of reality and being). By doing so, we hope to explore more fully issues of alignment related to teachers’ beliefs and practices. Therefore, we developed an integrated, holistic system in which teachers were asked to situate their epistemo­logical and ontological world views on scales ranging from realist to relativist perspectives. In this chapter, we report on the development

Lori Olafson and Gregory Schraw 517

of this two­dimensional scale and on the results of a study that utilized this scale. Twenty­four teachers rated their epistemological and onto­logical worldviews and then wrote an essay describing their ratings. Five of these teachers volunteered to participate in follow­up interviews when they were asked additional questions about their epistemologi­cal and ontological worldviews, the relationship between these beliefs, how their beliefs had changed over the course of their careers, and the impact of their beliefs on student learning.

The second issue is a methodological shortcoming in the type and quality of data collected that undermines a good understanding of how epistemological and ontological beliefs affect classroom practice. In their critique against quantification in educational theory, Guba and Lincoln (1994) describe two main issues with positivist research. One is context stripping, where variables within the context are overlooked. One important concern about context stripping is that it is difficult to understand the decisions and choices that teachers make in the class­room. The second problem is what Guba and Lincoln (1994) call the exclusion of meaning and purpose. They note that “human behavior, unlike that of physical objects, cannot be understood without regard to the meanings and purposes attached by human actors to their activities” (p. 106). While researchers may extrapolate and hypothesize regarding purposes, an alternative is to give teachers the opportunity to state, explain, and justify their purposes. Thus, in the present research, we endeavored to examine epistemological and ontological beliefs within the richer contextual framework in which teachers make decisions and select classroom practices based on local context and individual pur­poses. Our goal was to address both conceptual and methodological problems present in previous research in order to better understand classroom practices.

We believe that much of the previous research on epistemology, including our own, has fallen prey to both context stripping and excluding meaning and purpose. Gadamer (2002) notes that theo­retical constructions must be made fully good by experience. Our belief is that our conceptions of teachers’ epistemological beliefs can­not be “made fully good” by theorizing only about teachers’ epis­temological beliefs without attending to the conditions of teachers’ work and considering how teachers view themselves and their work in classrooms.

The current difficulties in trying to objectively explain “the way things really are” in terms of the relationship between teachers’ beliefs and practices perhaps stems from our attempt to erase the difficulties of everyday classroom life. A narrow focus on epistemology while

Assessing teachers’ worldviews518

researching teachers’ beliefs and practices occurs at the expense of developing an awareness of the ontological nature of teaching. Most research, and most techniques used in these studies to assess beliefs, confound epistemological and ontological beliefs.1 This is especially true within teacher education literature, where the unit of analysis is “teacher beliefs.” For example, in Speers’ (2005) study of two doctoral students teaching calculus courses, teacher beliefs included beliefs about the teacher’s role in the classroom; how students should solve problems; the domain of mathematics; the classroom atmosphere; and instructional practices such as collaborative work and questioning. It is essential for researchers to disentangle these beliefs as epistemo­logical (i.e., beliefs about the domain of mathematics) or ontological (i.e., beliefs about being a calculus teacher) in order to come to a bet­ter understanding of the complex relationship between them, and it is equally essential to understand these beliefs within the specific con­text of each teacher’s classroom. We conducted in­depth interviews in order to understand epistemological and ontological beliefs within the specific context and purposes of each teacher’s unique teaching situation.

Definitions

We introduce and define a number of terms in this section that we use regularly throughout the remainder of this chapter. These terms include epistemology, epistemological beliefs, epistemological worldviews, ontology, ontological beliefs, and ontological worldviews. Table 16.1 provides each term and its definition.

Epistemology is used in its broadest sense to refer to a theory of know­ledge and rationality (Lehrer, 1990; Pollock, 1986; Feldman, 2003). Hofer (2002, p. 4) defines epistemology as being “concerned with the origin, nature, limits, methods, and justification of human know­ledge.” Philosophical accounts of epistemology traditionally distinguish between kinds of knowledge (e.g., propositional or non­evidential) and justification of knowledge (e.g., weak or strong arguments) claims (Rescher, 2003). We use the term in this chapter to refer to the study of knowledge and beliefs about knowledge. We use this somewhat more restricted definition because most studies of student and teacher beliefs focus primarily on beliefs about knowledge rather than justification of knowledge claims.

1 See Schraw and Olafson (in press) for a more complete discussion of measurement issues related to epistemological and ontological beliefs.

Lori Olafson and Gregory Schraw 519

The term epistemological belief has been used widely for over a dec­ade to refer to a specific belief about some aspect of knowledge that is part of a broader epistemology. This implies that individuals may have more than one epistemological belief that is part of a set of beliefs that constitute a personal epistemology. Schommer (1990) proposed five independent beliefs pertaining to certain knowledge (i.e., abso­lute knowledge exists and will eventually be known), simple know­ledge (i.e., knowledge consists of discrete facts), omniscient authority (i.e., authorities have access to otherwise inaccessible knowledge), quick learning (i.e., learning occurs in a quick or not­at­all fashion), and innate ability (i.e., the ability to acquire knowledge is endowed at birth). Currently, there is debate as to whether Schommer’s five beliefs constitute genuine epistemological dimensions (Hofer and Pintrich, 1997; Pintrich, 2002). Most researchers agree that beliefs about the certainty and simplicity of knowledge constitute genuine epistemological beliefs. Hofer (2000) also has suggested the origin of knowledge as a genuine epistemological belief.

We use the term epistemological worldview to refer to an individual’s collective beliefs about the nature and acquisition of knowledge. These

Table 16.1. Key terms, definitions, and sources

Term Definition Recent sources

Epistemology The study of knowledge and beliefs about knowledge

Audi, 2003; Hofer, 2002; Feldman, 2003; Rescher, 2003

Epistemological beliefs Specific belief about some aspect of knowledge that is part of a broader epistemology (e.g., origin of knowledge)

Hofer, 2002; Schommer, 1990

Epistemological worldviews

A set of beliefs or theory about knowledge, acquisition of knowledge, and knowledge justification

Lincoln and Guba, 2000; Mertens, 2005; Schraw and Olafson, 2002

Ontology The study of the nature of reality and being

Mertens, 2005; Packer and Goicoechea, 2000; Merricks, 2007

Ontological beliefs A specific belief about some aspect of reality (e.g., realism)

Lincoln and Guba, 2000; Merricks, 2007; Shadish et al., 2002

Ontological worldviews

A set of beliefs or theory about reality or being (e.g., social constructivism)

Lincoln and Guba, 2000; Mertens, 2005; Ponterotto, 2005

Assessing teachers’ worldviews520

would include the five beliefs described by Schommer (1990) and the four belief taxonomy proposed by Hofer and Pintrich (1997). We use this term synonymously with similar terms used in the literature such as personal epistemology (Chan and Elliott, 2004; Hofer, 2001) and epis-temological stances (Johnston et al., 2001) that collectively refer to a set of beliefs or a personal theory about knowledge and knowledge jus­tification. We assume that an epistemological worldview includes all of one’s explicit and implicit beliefs, attitudes, and assumptions about the acquisition, structure, representation, and application of knowledge (Audi, 1995; Hofer and Pintrich, 1997; Pirttila­Backman and Kajanne, 2001; Schraw and Olafson, 2002).

It is important to distinguish clearly between epistemological beliefs and epistemological worldviews. The former consist of specific beliefs about a particular dimension of knowledge such as its certainty, sim­plicity, origin, or justification. The latter consist of a set of beliefs that collectively define one’s attitudes about the nature and acquisition of knowledge, and the justification of knowledge. Each adult presumably has a set of epistemological beliefs that are included within an epis­temological worldview, which also may include other beliefs such as how epistemological beliefs are acquired and develop, and how these beliefs change over time (Bendixen, 2002; Bendixen and Rule, 2004; Schommer­Aikins, 2002). Both epistemological beliefs and worldviews may be tacit or explicit (Feldman, 2003; Patrick and Pintrich, 2001; Schraw and Moshman, 1995). Presently, it is unclear whether tacit beliefs affect cognition and decision­making more or less than explicit beliefs, or whether they are easier or more difficult to change (How­ard et al., 2000; Sinatra and Pintrich, 2002; White, 2000). However, several studies indicate that epistemological beliefs and worldviews do not change quickly or easily. Typically, an individual experiences severe doubt and disequilibrium regarding a particular set of beliefs or world­view, and may return to his or her old belief system after exploring an alternative belief system (Ajzen, 1988; Asselin, 2000; Bendixen, 2002; Chandler, 1988; Chandler et al., 1990; Chandler et al., 2002; Richardson et al., 1991).

Ontology is used in its broadest sense to refer to the nature of reality and being (Lincoln and Guba, 2000; Merricks, 2007; Mertens, 2005; Ponterotto, 2005). Ontology addresses the question of “what is the form and nature of reality, and what can be known about that reality?” (Ponterotto, 2005, p. 130). Ontology usually is discussed independently of epistemology, but at some level, the two are related because beliefs about how we come to know reality necessarily involve epistemological assumptions. We assume that epistemological and ontological beliefs

Lori Olafson and Gregory Schraw 521

are related to some extent in most teachers’ minds, although this issue has not been examined empirically.

The term ontological belief has not been used widely in educational and psychological research, at least compared to the study of epistemo­logical beliefs. However, just as researchers have argued for the exist­ence of separate epistemological beliefs such as the origin or certainty of knowledge, it seems plausible that individuals would hold multiple ontological beliefs about the origin, permanence, and changeability of reality and being (Kuhn and Weinstock, 2002).

We use the term ontological worldview to refer to an individual’s col­lective beliefs about the nature of reality and being. We assume that an individual’s ontological beliefs collectively comprise a personal ontol­ogy. Like epistemology, ontological beliefs and worldviews may be tacit or explicit in part or whole. We also assume the epistemological and ontological worldviews work in tandem to determine an individual’s beliefs about learning and instruction. Indeed, many graduate textbooks in education research specifically discuss and compare methodological approaches that presume competing epistemological assumptions. For example, modes of research such as positivism, post­positivism, struc­turalism, and post­modernism assume mutually exclusive, incommen­surate assumptions about knowledge and reality (Cunningham and Fitzgerald, 1996; Lincoln and Guba, 2000; Mertens, 2005; Ponterotto, 2005; Shadish et al., 2002). Philosophers of science also have proposed theories of scientific change that draw on different epistemological and ontological underpinnings (Kuhn, 1962; Lakatos, 1978; Popper, 1959).

Previous research

Research over the last three decades has focused primarily on the struc­ture and development of college students’ epistemological beliefs ( Baxter Magolda, 1999, 2002; Hofer and Pintrich, 1997; King and Kitchener, 1994; Kuhn et al., 2000; Perry, 1970; Schommer, 1990). Research­ers have used one of six different quantitative measurement strategies for assessing epistemological and ontological beliefs and worldviews, including questionnaires (Hofer, 2000; Schommer, 1990; Schraw et al., 2002; Stahl and Bromme, in press), interviews (Johnston et al., 2001), vignettes (Joram, 2007; Schraw and Olafson, 2002), essays (Sandoval, 2003), and concept maps (Haerle, 2006; Ozgun­Koca and Sen, 2006). The vast majority of studies used questionnaires or interviews, with two or three studies using vignettes (Joram, 2007; Schraw and Olafson, 2002), essays (Brownlee, 2004; Olafson and Schraw, 2006; Sandoval,

Assessing teachers’ worldviews522

2003), and concept maps (Haerle, 2006; Liu, 2004; Ozgun­Koca and Sen, 2006). Many studies have used two or more of the six measure­ment strategies, typically with questionnaires or interviews as the pri­mary data collection strategy.

Researchers have begun investigating teachers’ epistemological beliefs in the last decade. Many of these researchers have argued that teach­ers’ epistemological beliefs influence teaching practices (Brownlee and Berthelsen, 2006; Chan and Elliott, 2004; Haney and McArthur, 2002; Ozgun­Koca and Sen, 2006; Trumbull et al., 2006; Yang, 2005) and that teacher beliefs may also have an impact on student epistemologi­cal development and learning (Johnston et al., 2001; Lidar et al., 2005; Louca et al., 2004; Marra, 2005). Discrepancies between teachers’ espoused beliefs and their enacted beliefs are commonly noted in the research. In particular, it seems that although preservice teachers indi­cate a preference for constructivist beliefs, they find it difficult to put these beliefs into practice (Haney and McArther, 2002; Ozgun­Koca and Sen, 2006).

The development of teachers’ epistemological beliefs is another area of study within the field. Brownlee (2004), for example, found that preservice teachers who were enrolled in a program based on rela­tional pedagogy experienced more growth in sophisticated epistemo­logical beliefs as compared with preservice teachers in a tutorial group. Marra (2005) reported similar findings in a study of how constructivist instruction affected the development of graduate student teachers at a university. Teachers reported a variety of changes after the course, but especially changes in epistemological pedagogical beliefs. Teach­ers adopted constructivist beliefs that emphasized the role of student interactions. In addition, graduate student teachers adopted stronger contextualist beliefs about the complexity and certainty of knowledge.

In contrast, little theoretical or empirical research has been con­ducted on students’ or teachers’ ontological beliefs by educational and psychological researchers. Ontology is mentioned in Rosenberg et al.’s (2006) study of student epistemologies in an eighth­grade science class. Their use of the term, however, is in reference to researchers’ concep­tualizations of students’ epistemologies and not to “ways of being” for teachers or students.

Slotta and Chi (2006) investigated the effect of ontological training in promoting conceptual change among physics students. Ontological training consisted of a computerized instructional module that helped students understand four ontological principles related to emergent processes in physical phenomena, which focused on changes that are system wide, equilibrium seeking, simultaneous and independent,

Lori Olafson and Gregory Schraw 523

and ongoing. College students were classified based on interviews into expert and novice levels of ontological commitment. Results indicated that experts performed better than novices on the post­test as well as the completeness of verbal explanations. The experimental group also outperformed the control on the posttest and verbal explanations. These findings indicated that explicit training helps students become aware of implicit ontological categories and to learn information and evaluation skills better than the control group.

Although there are few empirical studies investigating student and teacher ontology, recent review and analysis articles by Packer and Goicoechea (2000), Ponterotto (2005), and Kincheloe (2003) showed interest in pursuing the topic. Packer and Goicoechea (2000) called for the reintroduction of ontology as a topic in research on learning and development. The ontological processes of schooling described by Packer and Goicoechea (2000) illustrate how children are actively engaged in the ongoing reproduction of the classroom community of practice and how schools operate as sites for the production of persons. They noted that “the shift from family member to student is already an ontological transformation” (p. 235) as it fundamentally changes the “being” of the child.

Kincheloe (2003) developed a notion of critical ontology for teachers. A vision of critical ontology demands that teachers become political agents who research their own practices and their own belief sys­tems: “in doing so they develop their own teacher persona” (p. 53). The ways that teachers see themselves can then become connected to the social, political, cultural, economic, and historical world around them.

Instrument development

Thus far, self­report instruments that have been developed to assess multiple epistemological beliefs have had mixed success: findings have been inconsistent and measurement problems have been a persistent problem for researchers (Pintrich, 2002; Wood and Kardash, 2002). We are not aware of any instruments that have been developed to assess ontological beliefs. Therefore, we were interested in developing a new strategy for assessing epistemological and ontological worldviews using a common measurement scale. We refer to this as the four-quadrant scale because there are four distinct quadrants into which a person can be classified based on self­report or external judgment by a researcher. This approach is an application of the issues discussed by Shadish et al. (2002), regarding differences in individual beliefs about the theory and conduct of social science research. While Shadish et al. (2002) did not

Assessing teachers’ worldviews524

develop a measurement system, their ideas suggested a multidimen­sional system similar to the system we developed.

In reading a number of articles on perspectives on educational research, we were struck by the fact that authors deliberately separated epistemological issues from ontological issues when evaluating quanti­tative and qualitative research paradigms to help identify general meta­scientific principles (Guba and Lincoln, 1994; Maxwell, 2004; Shadish et al., 2002). Shadish et al. (2002) suggested that practicing scientists hold distinct epistemological and ontological beliefs that exist on a con­tinuum that ranges from realist to relativist endpoints. A realist believes that entities or phenomena (e.g., knowledge or physical matter) exist and can be understood and explained to some degree, even if experts do not currently understand the phenomenon that is being considered. For example, a physicist may believe that “dark matter” exists in open space even though it is currently undetectable. The basis for their belief may be theoretical (e.g., mathematical models), partial empirical evi­dence, or faith. In contrast, a relativist believes that entities may exist in an ever­changing manner (e.g., the changing nature of human rights), or that we can never know with certainty whether something exists (e.g., that God exists).

The assumptions made by Shadish et al. (2002) that epistemological and ontological beliefs are separate from one another and range from realist to relativist ends of a continuum is shared by many philoso­phers (Feldman, 2003; Mertens, 2005; Rescher, 2003). Our goal was to develop a measurement system that would allow us to collect data regarding the extent to which an individual endorsed a realist versus relativist view on each dimension. As a result, we developed a two­ dimensional framework as a tool for understanding teachers’ beliefs and practices (see Figure 16.1) based on the broad foundational distinctions between epistemology and ontology.

The four­quadrant scale partitions epistemological and ontological worldviews into two axes at right angles to each other that range from realist to relativist on each axis. This yields four quadrants in which a person can rate oneself as realist­realist, realist­relativist, relativist­ realist, or relativist­relativist. Individuals are able to select a point in any area of the four­quadrant array that best corresponds to their per­sonal epistemological and ontological worldviews about teaching.

The four labels indicating epistemological and ontological realism/relativism are included as anchor points in the current study. The figure used in the study used axes of equal length. We used a scale with two­right angle axes of 150 millimeter length (i.e., approximately six inches in length). The location of each participant’s X can be scored on a scale

Lori Olafson and Gregory Schraw 525

of 1–150 using a ruler scaled in millimeters. For example, an X in the extreme upper right corner would be scored as a 150 on the epistemo­logical axis and a 150 on the ontological axis. In contrast, an X at the intersection point of the two axes would be scored as a 75 on each dimen­sion. This enables researchers to assess both types of worldviews on the same scale. We assume that scores further from the point of origin are prototypical of a strong commitment to a worldview. In subsequent data collection these scores have been correlated to other measures, such as self­efficacy. For the purposes of the current study, given the small num­ber of participants, we do not report on individual epistemological and ontological scores. Rather, we focus on the participant’s point of selection on the four quadrant array as an overall indication of worldview.

Along the epistemological continuum, teachers’ personal epistemolo­gies can be located between the two endpoints of epistemological realist and epistemological relativist. An epistemological realist would believe that there is an objective body of knowledge. From a teacher’s perspec­tive, this position would hold that curriculum is fixed and permanent and focuses on fact­based subject matter. An epistemological relativist, on the other hand, would describe curriculum as changing and stu­dent­centered. Problem­based or inquiry curricula are examples at the epistemological relativist end of the continuum, while the perspective of a one­size­fits­all curriculum is at the epistemological realist end of the continuum.

Quadrant 4 Quadrant 1

Quadrant 3 Quadrant 2

Epistemologicalrelativist

Epistemologicalrealist

Ontologicalrelativist

Ontologicalrealist

Figure 16.1: The four­quadrant scale

Assessing teachers’ worldviews526

Ontology is the study of beliefs about the nature of reality and the nature of being: “What is, what exists, what it means for something – or somebody – to be” (Packer and Goicoechea, 2000). The who that is teaching (Field and Latta, 2001) or the “mode of being” (Foucault, 1987, p. 2) a teacher in the classroom is the way that we characterize the ontology of teachers (i.e., what does it mean to be a teacher?). A teach­er’s mode of being is related to instructional practices. For example, a teacher who views herself as an information­deliverer will have dif­ferent instructional practices from a teacher who considers herself to be a facilitator of student learning. A teacher who is an ontological realist assumes one underlying reality that is the same for everyone. Instructionally, this means that all children should receive the same type of instruction at the same time regardless of their individual cir­cumstances, achievement, or context. An ontological relativist, on the other hand, assumes that different people have different realities, and that these realities are constructed in social settings. From an instruc­tional perspective, teachers are seen as collaborators, co­participants, and facilitators of learning who work to meet the individual needs of students. Instructional practices are less teacher­directed. Uncovering the ways in which teachers view themselves and their students (their ontological beliefs about the classroom) are as important as describ­ing their personal epistemologies. From our perspective, teachers’ epis­temological and ontological beliefs both contribute to practices enacted in classrooms.

It should be noted that the two­dimensional framework in Figure 16.1 is a multidimensional scaling model that differs from unidimensional worldview taxonomy approaches (e.g., Kuhn, 1991; Perry, 1970) and multidimensional beliefs taxonomy approaches (e.g., Hofer, 2002; Schom­mer, 1990) in several important ways. One difference is that the Kuh­nian approach classifies an individual into one of several mutually exclusive categories (e.g., absolutist, multiplist, or evaluatist). This could be misleading in our opinion because individuals hold complex beliefs that may span multiple perspectives or classification categories. In contrast, the four­quadrant scale assigns an individual to a quad­rant based on a unique configuration of epistemological and ontological worldviews. Individuals are assumed to differ within quadrants as well as between quadrants.

Thus far, there has been no direct comparison between the four quad­rant and Kuhnian approaches. We assume that individuals who endorse strong realist epistemologoical and ontological worldviews would also be likely to endorse absolutism. In contrast, those endorsing strong rela­tivism on both dimensions should endorse multiplism. We assume that

Lori Olafson and Gregory Schraw 527

evaluatists would report less extreme scores and be positioned some­where in the center of Figure 16.1. We assume these individuals would adopt world views that are neither strongly realist or relativist; however, future research is needed to examine this conjecture.

A second difference is that the Kuhnian system does not provide explicit information about the extent to which an individual endorses realist versus relativist worldviews on the two dimensions. Typically, individuals with strong realist perspectives are assigned to the absolutist category, whereas individuals with strong relativist beliefs are assigned to the multiplist category. In contrast, using the four­quadrant scale, two individuals could be in the same quadrant, but demonstrate differ­ent commitments to realism and relativism on the epistemological and ontological dimensions.

A third difference, and one that is extremely important in our opin­ion, is that the Kuhnian system (and all others as well) does not distin­guish between epistemological and ontological beliefs or worldviews. As a result, it is impossible to determine how the two belief systems are related to one another or to other teacher variables such as self­efficacy, pedagogical practices, etc.

The four­quadrant framework shown in Figure 16.1 also differs from multidimensional beliefs approaches in that the former attempts to understand an individual’s epistemological and ontological worldviews on a scale in terms of commitment to realism or relativism, whereas the latter focuses on specific beliefs. The four­quadrant framework does not consider separate beliefs in certainty or the simplicity of knowledge. While it is possible to map separate beliefs onto the four­quadrant scale, no one has done so at this point in time. We view the multidimensional beliefs approach and the four quadrant approach as two complemen­tary ways to understand personal beliefs and worldviews.

Methods

Participants

Twenty­four teachers participated in the study. All participants were enrolled in graduate level courses in curriculum and instruction at a large western university. Half the participants (n = 12) were enrolled in a master’s of education program and were participating in a course enti­tled “topics in literacy.” The other participants (n = 12) were enrolled in Ph.D. programs and were taking a class entitled “research on teaching and schooling.” The average age was 38.2 years (37.3 years of age in the master’s class and 38.6 years of age in the doctoral class). In both groups,

Assessing teachers’ worldviews528

83 percent of the participants were female. Twenty­two participants (92 percent) self­identified their ethnicity as “white,” matching the nation­ally emerging trend of a monocultural teaching force (Groulx, 2001).Teachers had between two and twenty­three years’ classroom teaching experience at a variety of grade levels, ranging from kindergarten to twelfth grade. Most participants from the master’s level class (75 per­cent) had less than six years teaching experience, and all participants from the doctoral level class had taught for more than five years.

Data collection

Data collection occurred in two phases. In Phase 1, participants com­pleted the four­quadrant scale during their regularly scheduled classes at the mid­semester point of the spring semester, 2006. In Phase 2, participants were selected from among those who volunteered for a follow­up interview.

During Phase 1 the four­quadrant scale was administered by the researchers. Participants read general instructions and summaries of epistemological and ontological realist and relativist positions, shown in Appendix 16.A. They next considered their own worldviews and then rated themselves on the four­quadrant scale shown in Figure 16.1 by placing an X in one of the four quadrants that best corresponded to their personal epistemological and ontological worldviews.

Participants were not allowed to discuss their worldviews with others prior to making their ratings; however, the remainder of their class was devoted to discussing and comparing worldviews after complet­ing their rating and justification. These discussions were led by the class instructor. Individuals read instructions and completed the rat­ing sheet. This required approximately five minutes. Individuals next received ten minutes to provide a written explanation of their epistemo­logical and ontological worldviews. At the end of the structured essay, participants were asked to indicate whether or not they would be inter­ested in volunteering for a follow­up interview.

Eight participants volunteered to be interviewed. Six participants were selected, based on an attempt to obtain representation from quadrants one and four. Participants were not selected from quadrants two or three for two reasons: no participants from quadrant three volunteered to be interviewed, and there were no participants located in quadrant two. Par­ticipants agreed to be interviewed on campus, most often either before or after class. One interviewee did not arrive for her interview and sub­sequent attempts to reschedule were not successful. In­depth interviews lasting an hour or more were thus conducted with four doctoral students (Jane, Sara, Lynn, and Wayne) and one master’s student (Matt) using

Lori Olafson and Gregory Schraw 529

the procedure and protocol shown in Appendix 16.B. These interviews, conducted by the researchers, were tape­recorded and transcribed.

Both the structured essays and the in­depth interviews reduced the “context stripping” and “exclusion of goals and purposes” referred to by Guba and Lincoln (1994). In these forms of data collection, teachers were specifically requested to address the context of their epistemo­logical and ontological worldviews by situating their beliefs within the classroom and discussing how the everyday realities of teaching were connected to these beliefs and to their goals.

Findings

Two forms of analyses were conducted. First, participants’ locations on the quadrant were analyzed. Next, all textual data (the structured essays and the interview transcripts) were entered as primary documents into Atlas.ti. Atlas.ti is a software program that facilitates many of the activities involved in qualitative data analysis and interpretation, but does not auto­mate these processes (Muhr, 2004). Rather, it supports coding according to a scheme of categories (Alexa and Zuell, 2000). It offers an exploratory, yet systematic, approach to qualitative analysis (Muhr, 2004).

Analysis: four-quadrant scale

On the two­dimensional scale, most respondents located themselves in quadrants one and four, with only two respondents in quadrant three, and none in quadrant two (see Figure 16.2). Twenty­two of the twenty­four participants (92 percent) rated themselves as ontological relativists of some degree. In contrast, ten of the twenty­four (45 percent) of respond­ents rated themselves as epistemological realists to some degree.

We observed a statistically significant positive relationship between the epistemological and ontological dimensions, r = .45, p < .05, using a two­tailed test. We computed this correlation by assigning two separ­ate scores to each teacher’s rating. One score was scaled from 1 to 150 on the epistemological dimension, while the second score was scaled from 1 to 150 on the ontological dimension. The finding of a significant positive correlation between the two suggests that realist beliefs on one dimension are associated with realist beliefs on the second dimension.

Analysis: structured essays and interviews

As indicated earlier, the transcribed interviews and structured essays were imported as text files into Atlas.ti to facilitate data analysis. Two main procedures are utilized in Atlas.ti: analysis at the textual level and

Assessing teachers’ worldviews530

analysis at the conceptual level. Textual level analysis involves review­ing all text files, selecting text segments known as “quotations,” and assigning codes (Muhr, 2004). These analysis techniques reflect the ideas used in grounded theory (Muhr, 2004) in which a central fea­ture is constant comparative analysis (Strauss and Corbin, 1994). In this approach to analysis, data is systematically analyzed and beginning understandings are then elaborated and modified as incoming data are played against them (Strauss and Corbin, 1994). In the current study, we began with a detailed, line by line analysis of the first interviews to generate initial codes (Strauss and Corbin, 1998) that were reflective of the topic under study. For example, the following comment by Wayne was coded as shifting beliefs:

I think that life experience goes a long way but in terms of building or changing beliefs that gets really tricky because I think that people really want to hold on to their beliefs and I think that is a difficult thing to change. (interview)

The initial list of codes was refined as we analyzed each subsequent interview. For example, four “code families” were developed that clus­tered together individual codes into subsets. One such family, factors affecting practice, consisted of eight individual codes, including funding, testing, and administration (see Appendix 16.C for the complete coding scheme). This coding scheme was then also applied to the structured essays.

Ontological relativistQuadrant 4 Quadrant 1

Quadrant 3 Quadrant 2

Ontological realist

Epistemologicalrealist

Epistemologicalrelativist

Figure 16.2: Visual results

Lori Olafson and Gregory Schraw 531

At the conceptual level of analysis, “network views” were used to group codes into larger themes. These “network views” represent and explore complex information in a graphical manner (Muhr, 2004). Five of these graphical representations were created: one for each of three quadrants, one related to conflicts experienced by teachers, and one related to the development of beliefs. These graphical representations characterize the major findings that emerged from the textual data, which are described in the next section. First, we outline key characteristics of teachers’ beliefs and practices within each of the quadrants before discussing findings related to conflicts and the development of beliefs.

Quadrant one: ontological relativist/epistemological relativist

Fourteen participants located themselves in quadrant one. Individu­als in this quadrant were characterized primarily by their overwhelm­ing support of student­centered practices that included constructivist approaches, group projects/presentations, student­directed learning, inquiry, and critical thinking. In addition, there was an awareness of the importance of individual differences:

I believe that learning is a community event and that learning takes place at different levels and different times for all students. Accommodations must be made for each student to maximize their individual learning experience. The teacher is a guide and encourages students to advocate for themselves as well as their peers. Students are not automatons and cannot be kept on a strict learn­ing level of conformity. (Shelley: master’s student, essay)

In addition to their student­centered instructional practices, these teach­ers demonstrated views of knowledge that could be considered more rela­tivist. For example, one participant discussed the creation of knowledge:

I believe also that knowledge is created between an interaction between the students themselves. As they talk to each other they are gaining insight from each other, so that interaction also creates understanding, knowledge, and meaning. (Sara: doctoral student, interview)

The relativist–relativist position can be summarized as one in which teachers engage in student­centered instructional practices that are consistent with their personal epistemology and the belief that curric­ulum is not fixed and permanent.

Quadrant two: epistemological relativist/ontological realist

There were no participants that identified with this position on the quadrant. This finding is not particularly surprising, given that it seems

Assessing teachers’ worldviews532

highly incongruent that a teacher would hold simultaneous beliefs in a student­centered and changing curriculum while also believing that students should receive the same type of instruction at the same time.

Quadrant three: ontological realist/epistemological realist

Two participants located themselves in quadrant three and they endorsed traditionalist views, which included support for teacher­cen­tered instruction of a universal curriculum based on core knowledge and skills. One participant wrote: “I tend to believe that there exists an objective body of knowledge that in order for students to become know­ledgeable in a particular content area, there are certain skills that must be mastered” (Lois: doctoral student). Instructionally, she said that this was accomplished through repeatable testing. Kyra, a master’s student, agreed with this worldview and wrote:

I would characterize myself as an epistemological realist more than any other option. This may be because I teach kindergarten and believe there is a core of knowledge that students must know in order to be successful students. These things include letter names/sounds, numbers, colors, etc.

Quadrant four: ontological relativist/epistemological realist

Eight participants identified themselves within this quadrant. This is a position that is characterized by a belief in a core body of knowledge that can be taught using student­centered methods. For example, one of the master’s students, Mona, wrote: “I believe that there is a core of knowl­edge that all students should know. However that knowledge base is taught should be in ways that are engaging and meaningful to students.”

Although two­thirds of the participants were aligned to either rela­tivist or realist worldviews with respect to both epistemology and ontol­ogy (quadrants one and three), within the fourth quadrant there is an apparent lack of alignment between teachers’ epistemological and onto­logical worldviews. None of these teachers expressed concern in their essays about where they located themselves on the four­quadrant scale. However, the doctoral level participants clearly articulated conflicts related to beliefs and practices during the interviews, and this theme is discussed in the next section of the findings.

Conflicts between beliefs and practices

A key finding that emerged from the qualitative analysis of the interview data was related to conflicts experienced by teachers as they encountered

Lori Olafson and Gregory Schraw 533

the complexities of practice in their daily lives in classrooms. These conflicts often occurred as a result of external pressures. We identified several factors that influenced classroom practice: (1) administration; (2) structured programs; (3) accountability for practice; (4) funding; (5) testing; (6) time; (7) school context; and (8) classroom context. Of these eight factors, those related to administrative decisions occurred with the most frequency.

Administrative decisions at the school and district level were con­cerned with mandated, structured programs and specific instructional practices, testing, and accountability for practice. Most commonly, the structured programs and instructional practices that were mandated were firmly within traditional views of teacher­centered instruction, such as basal reading programs, and were viewed as a hindrance to practice. For example, Jane said: “Top­down administrators can be barrier to practice. Micromanaging everything. You must round robin read.”2

In the large school district where participants taught, there seemed to be a focus on structured programs as described by Sara:

The region that I was in they had adopted so many things like the reading series, the math series, even a self­esteem kind of program that was controlled by a set of curriculum and almost scripted program. So I think that knowing that the principal was expecting those they stepped away from that type of teaching.

This example shows the clear link between programs mandated by the principal and teacher­centered instruction that occurs in classrooms. “That type of teaching” referred to by the participant was reflective of more student­centered practices.

Site­based administrators can also require lesson plans to be written weeks in advance in order to provide evidence of accountability. As Jane noted, this is a practice that denies diversity: “I can’t imagine if I had to write out everything to differentiate my instruction or to meet diverse students’ needs. In my lesson plans pretty much what they want is the power standards and how you are assessing it. That is what it has come down to. Are you meeting the benchmarks? So it is not about whether anyone learned anything meaningful.”

On the other hand, administrators also may have a beneficial impact on practice. Three participants discussed ways in which administrators

2 Round robin reading is a whole­class oral reading strategy where each student takes a turn reading from a common text while the other students listen.

Assessing teachers’ worldviews534

were supportive of student­centered learning. For example, Sara ( doctoral student) said:

The way that I taught was based on the fact that my administrator allowed me to teach like that. They were not there to make sure that I was reading from a basal text or to teaching from the math series. They were there to make sure that the kids were learning so in that situation I wasn’t controlled by the cur­riculum or the administration.

Other external influences that had an impact on teacher practice included assessment programs in the schools (e.g., Sara: “And I think that unfortunately the fear that the test scores will go down will stop them from trying inquiry­based learning”), and lack of time. Partici­pants spoke of the pressure to “hit the syllabus” and teach to all the standards within a limited amount of time. Given time constraints, teacher­centered practices were viewed as the most efficient means of teaching to the tests. Also noted as an external influence to practice was the school context. Matt said, “The context where you teach can have an impact on practice,” and in particular he discussed having diffi­culty with parents at his new school. External influences such as these, often beyond the control of individual teachers, significantly affected instructional decisions within classrooms.

Development of beliefs

We were particularly interested in exploring how teachers developed beliefs and practices that were consistent with the relativist world­view. This group interested us because the relativist worldview may be more sophisticated developmentally, and they may be more open to alternative sources of evidence. Although more research needs to be conducted related to these two conjectures, they have important implications for teachers’ work in classrooms. During the interviews, participants were asked to articulate how their beliefs had shifted over the course of their teaching careers. Participants discussed experi­ences that precipitated these changes. We found that a variety of for­mal and informal experiences were noteworthy, including preservice and graduate course work, mentoring and collaboration at the school site, professional development opportunities, and personal reflection. These shifting beliefs were then associated with changing classroom practices. For example, Lynn said: “I think that the knowledge under­lying my practice has changed and so my practice itself would defi­nitely be more focused on students using what they already know to gain more knowledge.”

Lori Olafson and Gregory Schraw 535

Change in beliefs can be precipitated by a state of disequilibrium, as previous research has shown (Bendixen, 2002; Hofer and Pintrich, 1997; Murphy and Mason, 2006). Participants corroborated this notion, say­ing, for example, “I know that for me I need to be a little rattled before I am ready to shift my beliefs” (Lynn: doctoral student, interview). On the other hand, trying new instructional approaches could also occur prior to a shift in beliefs. Jane, another doctoral student, spoke about her experiences as a novice teacher when one of her colleagues placed research based­articles about reading practice in her mailbox:

The articles were along the lines of here is a different way to do evaluative reading or interactive writing. There are reasons why we do it this way. I never thought of that. I would see something done and think that is a good idea. I never thought about it as being, you know there was a purpose behind it.

In this example, Jane said that first she changed her practice, and then began to reflect on the changes.

Graduate level course work seemed to have a major impact on the ways that teachers viewed their work in classrooms. In particular, doctoral level participants clearly articulated the alignment of their epistemological and ontological worldviews. The majority of par­ticipants (75 percent) enrolled in the doctoral level class positioned themselves in quadrant one (ontological relativist/epistemological relativist), whereas the participants in the masters class were more evenly divided between quadrant one (42 percent as ontological rela­tivist/epistemological relativist) and quadrant four (50 percent as epistemological realist/ontological relativist). When asked to discuss the differences between the two groups, the doctoral level partici­pants suggested that masters level courses were focused primarily on instructional approaches in the classroom with little attention paid to the theoretical underpinnings of instruction. These results are similar to previous findings.

For example, Wilson (2000) used interviews and surveys to examine whether the amount of education was related to teachers’ epistemo­logical stances. Teachers were classified as dualist (i.e., belief that there is one right authority), multiplist (i.e., belief that different viewpoints exist), or relativist (i.e., belief that knowledge and values are context­ual). Teachers with graduate degrees tended to be in the multiplist and relativist groups: they preferred a student­centered classroom charac­terized by creativity, problem­solving, thinking about thinking, diverse viewpoints, interaction, and self­discovery (Wilson, 2000). In contrast, teachers with a baccalaureate degree or less were more likely to be in the dualist group: their classrooms were more structured and relied

Assessing teachers’ worldviews536

on lectures, concrete examples, rules, and little student participation (Wilson, 2000). Similarly, Joram (2007) concluded that students and beginning teachers hold a view of educational knowledge as noncu­mulative, specific, transmitted, and difficult to falsify; whereas experi­enced teachers and professors hold a view of educational knowledge as cumulative, changeable, generalizable, and falsifiable.

Discussion

In the current study, participants were able to complete the four­quad­rant scale quickly and efficiently. Individuals reported that the instruc­tions were easy to understand and that they had a good idea of what they were expected to do. Examining their positions on the scale and analyz­ing the structured essays yielded rich information that demonstrates the complexity of understanding the relationship between teachers’ beliefs and practices. A number of issues related to this relationship are dis­cussed next, including: (1) the impact of beliefs on instruction; (2) the alignment of beliefs and practices; (3) the development of beliefs; and (4) the extent to which teachers’ beliefs are related to student engage­ment and achievement.

The impact of beliefs on instruction

Findings from the current study suggested that epistemological and ontological beliefs impact teachers’ curricular and instructional choices. For example, previous research suggests that teachers with high ver­sus low self­efficacy adopt different instructional and classroom man­agement strategies (Calderhead, 1996; Goddard et al., 2000; Pajares, 1996; Woolfolk­Hoy and Burke­Spero, 2005). Several recent stud­ies also suggest that different epistemological beliefs lead to different teaching practices (Brownlee and Berthelsen, 2006; Chan and Elliott, 2004; Tsai, 2007; Yang, 2005). Teachers with more sophisticated epis­temological beliefs and worldviews were more likely to endorse student­centered instructional practices that emphasizes critical reasoning. Teachers with less sophisticated beliefs focused on traditional curric­ulum, student testing, and mastery of basic science concepts. Kang and Wallace (2004) reported similar findings when examining the relationship between epistemological beliefs and laboratory activities in American high school science classrooms. Lidar et al. (2005) found that teachers with more sophisticated personal epistemologies used a greater number of epistemological moves in their science classrooms, where moves consisted of cognitive activities designed to promote

Lori Olafson and Gregory Schraw 537

deeper learning and reflection, including generating, constructing, and re­constructing knowledge. They also reported that the relative success of different epistemological moves depended in large part on context­ually specific factors such as student knowledge, complexity of activity, and sophistication of students’ conceptual understanding. We found that teachers with strong relativist worldviews on either of the epis­temological or ontological dimensions are more likely to conduct a stu­dent­centered classroom (Brownlee and Berthelsen, 2006; Hashweh, 1996). However, at this time, it is unclear to us how epistemological and ontological worldviews are related to one another with respect to teach­ers’ instructional practices. Future studies should attempt to examine whether the two worldviews make separate contributions, or if one is more important than the other.

Alignment of beliefs and practices

Most participants were consistent with their epistemological and onto­logical worldviews (those located in quadrants one and three). Never­theless, more research needs to be conducted on this issue, especially with teachers who identified themselves as epistemological realists and ontological relativists (quadrant four).

One explanation for this apparent lack of consistency between epis­temological and ontological worldviews among some teachers is that within the current teaching climate, instructional decisions have been stripped away from teachers, leading to feelings of powerlessness, as Sara described:

I think unfortunately this district makes them want to teach programs and not teach children. I know that the teachers in the school that I was at struggled with that. I don’t think that they wanted to teach all of those programs – they even complained about all of those programs – but I don’t think they could step away from it because of all the control the district and administration had.

In their study of urban middle school teachers, Shulman and Armit­age (2005) found that many teachers let existing curriculum guide what they do in the classroom: “Most of them plan instruction around curriculum handed to them by their supervisors” (p. 387). Given that the content is predetermined and unchanging, it seems reasonable to expect that teachers within this type of environment are epistemological realists. One of the key differences between the relativists and realists in the current study was that relativists did not view the mandated cur­riculum as a barrier to their practice. Instead, curriculum was viewed

Assessing teachers’ worldviews538

as a roadmap: “I saw the curriculum as a roadmap and I go that direc­tion but I could take any route that I need to get there” (Wayne: inter­view). Another participant, Sara remarked during her interview that “There are ways around the mandated curriculum” and she provided an example of how this was accomplished in her classroom. Realists, on the other hand, appear to be willing to unquestioningly accept man­dated curriculum.

In their study of secondary school art teachers, Lam and Kem­ber (2006) maintain that as contextual inf luences grow, they start to have greater inf luence on the way teachers teach. For example, pressures to succeed on state standardized tests, endorse a particular curriculum, or maintain existing performance levels, may lead to a divorce between teacher beliefs and practice. In light of our results, relativists and realists may assign contextual inf luences greater or lesser roles in their daily teaching practice. Furthermore, studies by Kember and Kwan (2000) and Trigwell and Prosser (1996a, 1996b) show that teachers in higher education may have limited external inf luences on what and how they teach compared to the public sector. For example, academic freedom, internal assessment, self­accredita­tion courses, and relatively unobtrusive quality assurance procedures may combine to allow teachers in higher education greater latitude. The result may be closer alignment of epistemological and ontologi­cal beliefs and practice.

The majority of epistemological realists (80 percent) also endorsed an ontological relativist view. As mentioned previously, these par­ticipants did not indicate any concern about a possible misalignment between their ontological and epistemological worldviews. There could be a number of reasons for individuals choosing quadrant four, ran­ging from lack of meta­cognitive awareness of the inconsistency in one’s beliefs, to a plausible justification of how the different beliefs can be reconciled in a credible way.

Another possible explanation comes from Slotta and Chi’s (2006) study of students learning physics. According to Slotta and Chi (2006), students with misattributed ontology may need to revise their onto­logical commitments. With respect to teachers, it may be the case that they revise their ontological commitments (i.e., a belief about the socially constructed reality of the classroom) when faced with some of the ontological processes of schooling that interfere with their capacity to practice their beliefs. Being a “good teacher,” for example, might be defined by the principal as one who closely follows a scripted reading program, which could be at odds with the teacher’s beliefs about meet­ing students’ individual reading needs.

Lori Olafson and Gregory Schraw 539

Development of beliefs

A third issue concerns teacher development over time (Garet et al., 2001; Kuhn et al., 2000). Previous research suggests that teachers develop over time regarding teaching practices and that their beliefs change as part of their development (Brownlee et al., 2001; Gill et al., 2004; Woolfolk et al., 2006). We wish to elaborate on two assump­tions made by Shadish et al. (2002) that we concur with. One is that beliefs occur on a continuum and may change over time from realist to relativist or the reverse. A second is that the epistemological beliefs and worldviews held by an individual may be at one point on the continuum even though the same individual’s ontological beliefs and worldview may be at a different point on the continuum. Thus, a person’s com­mitment to a realist or relativist point of view may change over time and differ across epistemological and ontological dimensions. We assume that beliefs are changeable due to a variety of factors, but espe­cially education, explicit inquiry, collaborative discussion, and devel­oping critical reasoning skills (Kuhn, 1999; White and Frederiksen, 2005). Our results are consistent with this research, and we identified a number of formal and informal experiences, such as graduate course work, mentoring, and collaboration, that were related to teacher devel­opment. However, very little is known currently regarding the devel­opment of teachers’ epistemological and ontological beliefs (Bendixen, 2002). One question is whether the two beliefs develop in tandem or follow separate trajectories. A second question is the general trend in development. We assume that beliefs become more relativist over time, but may become stable at some point, or perhaps move from a relativ­ist to realist worldviews past a certain number of years of experience (Lieberman, 1995; Reybold, 2001). The majority of doctoral level par­ticipants in the current study endorsed relativist worldviews. A com­bination of advanced course work and years of teaching experience seemed to lead to these views.

Relationship of beliefs to student engagement and achievement

A final issue is the extent to which teachers’ epistemological and onto­logical beliefs are related to student engagement and achievement. We assume that relativist teachers conduct a more constructivist­oriented classroom that is more likely to engage students and promote deeper learn­ing (Holt­Reynolds, 2000). A learner­centered teaching style and teach­ers who believe in the ability of all students to learn is also associated with

Assessing teachers’ worldviews540

improved student achievement (Opdenakker and Van Damme, 2006). Participants who were interviewed in the current study provided exam­ples of how teacher beliefs affected student learning. Lynn, for example, thought that students were less likely to shut down in a classroom when they knew that more than one answer could be correct:

For kids in a classroom that are faced with a teacher who is a realist, I think it can shut them down. Part of that is coming from my experience working with the gay/straight alliance and talking to kids who felt that some of their teachers were not willing to hear where they were coming from, and they would not turn in an essay in English class. So it really hurt their performance in class and it made those classes difficult to attend. (interview)

Jane also agreed in her interview that teacher beliefs affected student learning, and provided an example from her own practice:

I think because I was open to them wanting to learn new things and I believed that they could learn things without my controlling them, I think they under­stood that if there was something really exciting they wanted to learn that we would make time for that. I think that me saying “No, it doesn’t have to be knowing everything, it is us working together” made them feel comfortable enough to help me teach them.

Collectively, our findings support the view that teachers’ personal epistemology in the classroom matters and that epistemological and ontological worldviews impact instruction and student learning, even though teachers also may engage in instructional practices that may have little relationship to their worldviews. We identified a number of external influences that had profound impacts on instruction. Con­textual factors such as authorities imposing restrictions on curriculum are related to teachers becoming more controlling with their students ( Pelletier et al., 2002), as well as demands for preparation and comple­tion of norm­referenced assessments.

Limitations and future research

The present findings should be viewed as exploratory. We caution that the current study was our first attempt at piloting the multidimensional scaling system known as the four­quadrant scale. However, we have subsequently collected data using this scale with more than 300 partici­pants. As a result, we have been refining and revising the vignettes and the directions for completion.

We believe there are at least three distinct measurement systems that researchers can use to assess different aspects of personal epistemology

Lori Olafson and Gregory Schraw 541

and ontology. These include the Kuhnian approach, in which individ­uals are classified into one of three mutually exclusive categories; the multidimensional beliefs approach, which examines the role of individ­ual epistemological beliefs; and the four­quadrant scale, which situates an individual on a two­dimensional scale with respect to their epis­temological and ontological worldviews. From a measurement stand­point, these systems are not mutually exclusive; thus, we encourage researchers to examine the relationships across these measurement systems to better understand individuals, as well as the relationships between epistemological beliefs and worldviews, and epistemological and ontological worldviews. We also encourage researchers to inquire into the dimensionality of ontological beliefs. Previous research has argued for five (Schommer, 1990) or four (Hofer and Pintrich, 1997) core epistemological dimensions. It is possible that multiple ontological dimensions exist as well.

In addition, future research is needed to explore in greater detail the development of teachers’ epistemological and ontological beliefs, as well as the relationships between teacher and student beliefs. Exploring the day­to­day decisions and choices that teachers make in the classroom would be extremely informative. More research is needed to under­stand as well the alignment between stated beliefs versus practices that are inconsistent with those beliefs. Discrepancies between beliefs and practice seem to be common and may affect teacher efficacy and burn out. In extreme cases, teachers such as Sara may elect to leave the teaching profession when it becomes impossible to reconcile differences between beliefs and practices:

We had to give grades in library. And so that actually made me leave the school district because I felt like now that I had gone from being able to let the students choose and inquiry­based learning and went to I need to have these types of assessment and it had to fit. That was really hard for me so I ended up leaving.

R EF ER ENCES

Ajzen, I. (1988). Attitudes, personality, and behavior. Milton Keynes, UK: Open University Press.

Alexa, M. and Zuell, C. (2000). Text analysis software: Commonalities, dif­ferences and limitations: The results of a review. Quality and Quantity, 34, 299–321.

Asselin, M. (2000). Confronting assumptions: Preservice teachers’ beliefs about reading and literature. Reading Psychology, 21, 31–55.

Audi, R. (1995). The Cambridge dictionary of philosophy. Cambridge: Cambridge University Press.

Assessing teachers’ worldviews542

(2003). Epistemology: A contemporary introduction to the theory of knowledge (2nd edn.). New York: Routledge.

Baxter Magolda, M. B. (1999). The evolution of epistemology: Refining con­textual knowing at twentysomething. Journal of College Student Develop-ment, 36, 205–16.

(2002). Epistemological reflection: The evolution of epistemological assump­tions from age 18 to 30. In B. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (89–102). Mahwah, NJ: Lawrence Erlbaum Associates.

Bendixen, L. D. (2002). A process model of epistemic belief change. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (191–208). Mahwah, NJ: Lawrence Erlbaum Associates.

Bendixen, L. D. and Rule, D. C. (2004). An integrative approach to personal epistemology: A guiding model. Educational Psychologist, 39, 69–80.

Brownlee, J. (2004). Teacher education students’ epistemological beliefs. Research in Education, 72, 1–17.

Brownlee, J. and Berthelsen, D. (2006). Personal epistemology and relational pedagogy in early childhood teacher education programs. Early Years, 26, 17–29.

Brownlee, J., Purdie, N., and Boulton­Lewis, G. (2001). Changing episte­mological beliefs in pre­service teaching education students. Teaching in Higher Education, 6, 247–68.

Calderhead, J. (1996). Teachers: Beliefs and knowledge. In D. C. Berliner and R. C. Calfee (Eds.), Handbook of educational psychology (709–25). New York: MacMillan.

Chan, K. and Elliott, R. G. (2004). Relational analysis of personal epistemol­ogy and conceptions about teaching and learning. Teaching and Teacher Education, 20, 817–31.

Chandler, M. (1988). Doubts and developing theories of mind. In J. W. Astington et al. (Eds.), Developing theories of mind (387–413). Cam­bridge: Cambridge University Press.

Chandler, M., Boyes, M., and Ball, L. (1990). Relativism and stations of epis­temic doubt. Journal of Experimental Child Psychology, 50, 370–95.

Chandler, M. J., Hallert, D., and Sokol, B. W. (2002). Competing claims about competing knowledge claims. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (145–67). Mahwah, NJ: Lawrence Erlbaum Associates.

Cunningham, J. W. and Fitzgerald, J. (1996). Epistemology and reading. Read-ing Research Quarterly, 31, 36–60.

Feldman, R. (2003). Epistemology. Upper Saddle River, NJ: Prentice Hall.Field, J. and Latta, M. (2001). What constitutes becoming experienced in

teaching and learning? Teaching and Teacher Education, 17, 885–95.Foucault, M. (1987). The ethic of care for the self as a practice of freedom. In

J. Bernauer and D. Rasmussen (Eds.), The final Foucault. Cambridge: The MIT Press.

Gadamer, H. G. (2002). Question and answer play back and forth between the text and its interpreter. In D. R. Steele (Ed.), Genius: In their own words. Chicago, IL: Open Court.

Lori Olafson and Gregory Schraw 543

Garet, M. S., Porter, A. C., Desimone, L., Birman, B. F., and Yoon, K. S. (2001). What makes professional development effective? Results from a national sample of teachers. American Educational Research Journal, 38, 915–45.

Gill, M. G., Ashton, P. T., and Algina, J. (2004). Changing preservice teachers’ epistemological beliefs about teaching and learning in mathematics: An intervention study. Contemporary Educational Psychology, 29, 164–85.

Goddard, R. D., Hoy, W. K., and Woolfolk Hoy, A. (2000). Collective teacher efficacy: Its meaning, measure and impact on student achievement. Amer-ican Educational Research Journal, 37, 479–507.

Groulx, J. (2001). Changing preservice teacher perceptions of minority schools. Urban Education, 36(1), 60–92.

Guba, E. and Lincoln, Y. (1994). Competing paradigms in qualitative research. In N. Denzin and Y. Lincoln (Eds.), Handbook of qualitative research (105–17). Thousand Oaks, CA: Sage.

Haerle, F. C. (2006). Personal epistemologies of fourth graders: Their beliefs about knowledge and knowing. Oldenburg: Didaktisches Zentrum.

Haney, J. and McArthur, J. (2002). Four case studies of prospective science teachers’ beliefs concerning constructivist practices. Science Education, 86, 783–802.

Hashweh, M. Z. (1996). Effects of science teachers’ epistemological beliefs in teaching. Journal of Research in Science Teaching, 33, 47–63.

Hofer, B. K. (2000). Dimensionality and disciplinary differences in personal epistemology. Contemporary Educational Psychology, 25, 378–405.

Hofer, B. (2001). Personal epistemology research: Implications for learning and teaching. Educational Psychology Review, 13, 353–84.

(2002). Epistemological world views of teachers: From beliefs to prac­tices. Issues in Education: Contributions from Educational Psychology, 8(2), 167–73.

Hofer, B. K. and Pintrich, P. R. (1997). The development of epistemological theories: Beliefs about knowledge and knowing and their relation to learn­ing. Review of Educational Research, 67, 88–140.

Holt­Reynolds, D. (2000). What does the teacher do? Constructivist pedago­gies and prospective teachers’ beliefs about the role of the teacher. Teach-ing and Teacher Education, 16, 21–32.

Howard, B. C., McGee, S., Schwartz, N., and Purcell, S. (2000). The expe­rience of constructivism: Transforming teacher epistemology. Journal of Research on Computing in Education, 32, 455–65.

Johnston, P., Woodside­Jiron, H., and Day, J. (2001). Teaching and learning literate epistemologies. Journal of Educational Psychology, 93(1), 223–33.

Joram, E. (2007). Clashing epistemologies: Aspiring teachers’, practicing teachers’, and professors’ beliefs and knowledge and research in educa­tion. Teaching and Teacher Education, 23, 123–35.

Kang, N. W. and Wallace, C. S. (2004). Secondary science teachers’ use of laboratory activities: Linking epistemological beliefs, goals, and practices. Science Education, 88, 140–65.

Kember, D. and Kwan, K. P. (2000). Lecturers’ approaches to teaching and their relationship to conceptions of good teaching. In W. Hativa and P. Goodyear (Eds.), Teacher thinking, beliefs, and knowledge in higher educa-tion. Dorderecht: Kluwer Academic Publishers.

Assessing teachers’ worldviews544

Kincheloe, J. (2003). Critical ontology: Visions of selfhood and curriculum. Journal of Curriculum Theorizing, 19(1), 47–64.

King, P. M. and Kitchener, K. S. (1994). Developing reflective judgment. San Francisco: Jossey­Bass.

Kuhn, D. (1991). The skills of argument. New York: Cambridge University Press.

(1999). A developmental model of critical thinking. Educational Researcher, 28, 16–26.

Kuhn, D., Cheney, R., and Weinstock, M. (2000). The development of episte­mological understanding. Cognitive Development, 15, 309–28.

Kuhn, D. and Weinstock, M. (2002). What is epistemological thinking and why does it matter? In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing (121–44). Mahwah, NJ: Lawrence Erlbaum Associates.

Kuhn, T. S. (1962). The structure of scientific revolutions. Chicago, IL: University of Chicago Press.

Lam, B. H. and Kember, D. (2006). The relationship between conceptions of teaching and approaches to teaching. Teachers and Teaching: Theory and Practice, 12, 693–713.

Lakatos, I. (1978). The methodology of scientific research programmes. Cambridge: Cambridge University Press.

Lehrer, K. (1990). Theory of knowledge. San Francisco, CA: Westview Books.Levitt, K. E. (2001). An analysis of elementary teachers’ beliefs regarding the

teaching and learning of science. Science Education, 86, 1–22.Lidar, M., Lundqvist, E., and Ostman, L. (2005). The interplay between

teachers’ epistemological moves and students’ practical epistemology. Sci-ence Education, 89, 149–63.

Lieberman, A. (1995). Practices that support teacher development: Trans­forming conceptions of professional development. Phi Delta Kappan, April, 591–96.

Lincoln, Y. S. and Guba, E. G. (2000). Paradigmatic controversies, contra­dictions, and emerging confluences. In N. K. Denzin and Y. S. Lincoln (Eds.), Handbook of qualitative research (2nd edn., 163–88). Thousand Oaks, CA: Sage Publications.

Louca, L., Elby, A., Hammer, D., and Kagey, T. (2004). Epistemological resources: Applying a new epistemological framework to science instruc­tion. Educational Psychologist, 39(1), 57–68.

Liu, X. (2004). Using conceptual mapping for assessing and promoting rela­tional conceptual changes in science. Science Education, 88, 373–96.

Marra, R. (2005). Teacher beliefs: The impact of the design of constructiv­ist learning environments on instructor epistemologies. Learning Environ-ments Research, 8, 135–55.

Merricks, T. (2007). Truth and ontology. Oxford: Oxford University Press.Mertens, D. M. (2005). Research and evaluation in education and psychology: Inte-

grating diversity with quantitative, qualitative, and mixed methods (2nd edn.). Thousand Oaks, CA: Sage Publications.

Maxwell, J. (2004). Causal explanation, qualitative research, and scientific inquiry in education. Educational Researcher, 33(2), 3–11.

Lori Olafson and Gregory Schraw 545

Muhr, T. (2004). ATLAS.ti v5.0 user’s guide and reference (2nd edn.). Berlin: Scientific Software Development, 417.

Murphy, P. K. and Mason, L. (2006). Changing knowledge and beliefs. In P. Alexander and P. Winne (Eds.), Handbook of educational psychology (2nd edn., 305–24). Mahwah, NJ: Lawrence Erlbaum Associates.

Olafson, L. J. and Schraw, G. (2002). Some final thoughts on the epistemologi­cal melting pot. Issues in Education, 8, 233–46.

(2006). Teachers’ beliefs and practices within and across domains. Interna-tional Journal of Educational Research, 45, 71–84.

Opdenakker, M. and Van Damme, J. (2006). Teacher characteristics and teaching styles as effectiveness enhancing factors of classroom practice. Teaching and Teacher Education, 22, 1–21.

Ozgun­Koca, S. A. and Sen, A. I. (2006). The beliefs and perceptions of pre­service teachers enrolled in a subject area dominant teacher education program about “Effective Education.” Teaching and Teacher Education, 22, 946–60.

Packer, M. and Goicoechea, J. (2000). Sociocultural and constructivist theo­ries of learning: Ontology, not just epistemology. Educational Psychologist, 35(4), 227–41.

Pajares, F. (1996). Self­efficacy beliefs in academic settings. Review of Educa-tional Research, 66, 543–78.

Patrick, H. and Pintrich, P. R. (2001). Conceptual change in teachers’ intui­tive conceptions of learning, motivation, and instructions: The role of motivational and epistemological beliefs. In B. Torf and R. Sternberg (Eds.), Understanding and teaching the intuitive mind (117–43). Mahwah, NJ: Lawrence Erlbaum Associates.

Pelletier, L., Seguin­Levesque, C., and Legault, L. (2002). Pressure from above and pressure from below as determinants of teachers’ motivation and teaching behaviors. Journal of Educational Psychology, 94(1), 186–96.

Perry, W. G. Jr. (1970). Forms of intellectual and ethical development in the college years. New York: Academic Press.

Pintrich, P. (2002). Future challenges and directions for theory and research on personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Per-sonal epistemology: The psychology of beliefs about knowledge and knowing (389–414). Mahwah, NJ: Lawrence Erlbaum Associates.

Pirttila­Backman, A. M. and Kajanne, A. (2001). The development of implicit epistemologies during early and middle adulthood. Journal of Adult Devel-opment, 8, 81–97.

Pollock, J. L. (1986). Contemporary theories of knowledge. Savage, Maryland: Rowman and Littlefield Publishers.

Ponterotto, J. G. (2005). Qualitative research in counseling psychology: A pri­mer on research paradigms and philosophy of science. Journal of Counsel-ing Psychology, 52, 126–36.

Popper, K. R. (1959). The logic of scientific discovery. London: Hutchinson.Rescher, N. (2003). Epistemology: An introduction to the theory of knowledge.

Albany, NY: State University of New York Press.Reybold, L. E. (2001). Encouraging the transformation of personal epistemol­

ogy. Qualitative Studies in Education, 14, 413–28.

Assessing teachers’ worldviews546

Richardson, V., Anders, P., Tidwell, D., and Lloyd, C. (1991). The relationship between teachers’ beliefs and practices in reading comprehension instruc­tion. American Educational Research Journal, 28, 559–86.

Rosenberg, S., Hammer, D., and Elby, A. (2006). Multiple epistemological coherences in an eighth­grade discussion of the rock cycle. Journal of the Learning Sciences, 15, 261–92.

Sandoval, W. A. (2003). Conceptual and methodological aspects of students’ scientific explanations. The Journal of the Learning Sciences, 12, 5–51.

Schommer, M. (1990). Effects of beliefs about the nature of knowledge on comprehension. Journal of Educational Psychology, 82, 498–504.

Schommer­Aikins, M. (2002). An evolving theoretical framework for an epis­temological belief system. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (103–18). Mahwah, NJ: Lawrence Erlbaum Associates.

Schraw, G. and Moshman, D. (1995). Metacognitive theories. Educational Psy-chology Review, 7, 351–71.

Schraw, G. and Olafson, L. (2002). Teachers’ epistemological world views and educational practices. Issues in Education, 8, 99–148.

(in press). Assessing teachers’ epistemological and ontological world views. In M. Khine (Ed.), Knowing, knowledge, and beliefs: Epistemological studies across diverse cultures.

Schraw, G., Bendixen, L. D., and Dunkle, M. E. (2002). Development and validation of the epistemic belief inventory (EBI). In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (261–75). Mahwah, NJ: Lawrence Erlbaum Associates.

Shadish, W. R., Cook, T. D., and Campbell, D. T. (2002). Experimental and quasi-experimental designs for generalized causal inference. Boston, MA: Houghton Mifflin Co.

Shulman, V. and Armitage, D. (2005). Project discovery: An urban middle school reform effort. Education and Urban Society, 37(4), 371–97.

Sinatra, G. M. and Pintrich, P. R. (Eds.) (2002). Intentional conceptual change. Mahwah, NJ: Lawrence Erlbaum Associates.

Slotta, J. D. and Chi, M. (2006). Helping children understand challenging topics in science through ontology training. Cognition and Instruction, 24, 261–89.

Speers, N. (2005). Issues of methods and theory in the study of mathemat­ics teachers’ professed and attributed beliefs. Educational Studies in Math-ematics, 58, 361–91.

Stahl, E. and Bromme, R. (in press). The CAEB: An instrument for measuring the connotative aspects of epistemological beliefs. Learning and Instruction.

Strauss, A. and Corbin, J. (1994). Grounded theory methodology: An over­view. In N. Denzin and Y. Lincoln (Eds.), Handbook of qualitative research. Thousand Oaks, CA: Sage.

(1998). Basics of qualitative research: Techniques and procedures for developing grounded theory. Thousand Oaks, CA: Sage.

Trigwell, K. and Prosser, M. (1996a). Changing approaches to teaching: A relational perspective. Studies in Higher Education, 21, 275–84.

Lori Olafson and Gregory Schraw 547

(1996b). Congruence between intention and strategy in university science teachers’ approaches to teaching. Higher Education, 32, 77–87.

Trumbull, D., Scarano, G., and Bonney, R. (2006). Relations among two teachers’ practices and beliefs, conceptualizations of the nature of science, and their implementation of student independent inquiry projects. Inter-national Journal of Science Education, 28(14), 1717–50.

Tsai, C. C. (2007). Teachers’ scientific epistemological views: The coherence with instruction and students’ views. Science Education, 92, 222–43.

White, B. C. (2000). Pre­service teachers’ epistemology viewed through per­spectives on problematic classroom situations. Journal of Education for Teaching, 26, 279–305.

White, B. and Frederiksen, J. (2005). A theoretical framework and approach for fostering metacognitive development. Educational Psychologist, 40, 211–23.

Wilson, B. A. (2000). The epistemological beliefs of technical college instruc­tors. Journal of Adult Development, 7, 179–86.

Wood, P. and Kardash, C. A. (2002). Critical elements in the design and analy­sis of studies of epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (261–75). Mahwah, NJ: Lawrence Erlbaum Associates.

Woolfolk, A., Winne, P. H., and Perry, N. E. (2006). Educational psychology: Can-adian edition (3rd edn.). Scarborough, Ontario: Allyn and Bacon, Canada.

Woolfolk­Hoy, A. and Burke­Spero, R. (2005). Changes in teacher efficacy during the early years of teaching: A comparison of four measures. Teach-ing and Teacher Education, 21, 457–71.

Yang, F. (2005). Student views concerning evidence and the expert in reason­ing a socio­scientific issue and personal epistemology. Educational Studies, 31, 65–84.

A PPEN DI X 16.AI NSTRUCT IONS TO PA RT ICIPA N TS

Instructions

We want you to rate and explain your epistemological and ontological world views. Please read the following description of terms used in this study. Then indicate with an “X” where you would place yourself in the four quadrants shown on the rating sheet. To make your X, find the point where your ratings intersect on the epistemology dimension and the ontology dimensions.

Please note that the descriptions provided below represent endpoints on each of the scales. Your own beliefs may lie anywhere between these two endpoints. You may use any part of the four quadrant area.

After you make your rating, please describe in as much detail as pos­sible on the explanation sheet your reasoning for your self­rating.

Assessing teachers’ worldviews548

Epistemology

Epistemology is the study of what can be counted as knowledge, where knowledge is located, and how knowledge increases. The personal epis­temology of teachers is characterized by a set of beliefs about learning and the acquisition of knowledge that drives classroom instruction.

Epistemological realist

An epistemological realist would believe that there is an objective body of knowledge that must be acquired. From a teacher’s perspective, this position would hold that curriculum is fixed and permanent and focuses on fact­based subject matter. An epistemological realist might believe the following:

There are certain things that students simply need to know.I am teaching information that requires memorization and

mastery.There are specific basic skills that need to be mastered.

Epistemological relativist

An epistemological relativist would describe curriculum as changing and student­centered. Problem­based or inquiry curricula are exam­ples at the other end of the continuum from a perspective of a one size fits all curriculum. One of the central features of curriculum from this position is the notion that curriculum is not fixed and permanent. An epistemological relativist might agree with the following statements:

The things we teach need to change along with the world.The content of the curriculum should be responsive to the

needs of the community.It is useful for students to engage in tasks in which there is no

indisputably correct answer.Students design their own problems to solve.

Ontology

Ontology is the study of beliefs about the nature of reality. The per­sonal ontology of teachers is characterized by a set of beliefs regarding whether students share a common reality and what a classroom reality should look like.

Lori Olafson and Gregory Schraw 549

Ontological realist

A teacher who is an ontological realist assumes one underlying real­ity that is the same for everyone. Instructionally, this means that all children should receive the same type of instruction at the same time regardless of their individual circumstances and context. An ontologi­cal realist would agree with the following:

Student assignments should always be done individually.It’s more practical to give the whole class the same assignment.The teacher must decide on what activities are to be done.

Ontological relativist

An ontological relativist assumes that different people have different realities. From an instructional perspective, teachers are seen as col­laborators, co­participants, and facilitators of learning who work to meet the individual needs of students. Instructional practices are less teacher­directed, such as:

Students need to be involved in actively learning through dis­cussions, projects, and presentations.

Students work together in small groups to complete an assign­ment as a team.

A PPEN DI X 16.BI N T ERV IEW PROM P TS

Teachers’ perceptions of their epistemological and ontological beliefs interview prompts

Provide the participant with his/her copy of the structured essay task and ask the following questions:

Discuss the reason for placing yourself in a particular quadrant.•What statements in the descriptions did you most strongly identify •with?What statements in the descriptions did you most strongly disagree •with?Which viewpoint did you most strongly disagree with?•Discuss specific examples in your classroom that are consistent with •these statements.

Assessing teachers’ worldviews550

What are classroom/school/school district factors that support prac­•ticing your beliefs?What are classroom/school/school district factors that act as barriers •to practicing your beliefs?What are some of the experiences that have shaped your current •beliefs about learning?Do a teacher’s beliefs affect student learning?•Discuss the relationship between your beliefs and practices in the •classroom.

Are your teaching practices reflective of your stated beliefs?•Provide specific examples.•

Have your beliefs and practices changed over your teaching career?•How? Why?•

A PPEN DI X 16.CCODI NG SCHEM E

Conflict (code family)

CB: conflict between beliefs and practice•CB: conflicting beliefs with others•CB: conflicting practices•CB: consequences of conflict•CB: negotiating between differences•

Changing practiceDeveloping beliefsEpistemological realist ontological realistEpistemological realist ontological relativistExperiences impacting beliefs (code family)

EB: grad classes•EB: life experiences•EB: mentoring•EB: observation•EB: professional development•EB: reflection•EB: teacher preparation•

Epistemological realismEpistemological relativismFactors affecting practice (code family)

FP: accountability for practice•FP: administration•

Lori Olafson and Gregory Schraw 551

FP: classroom context•FP: funding•FP: school context•FP: structured programs•FP: testing•FP: time•

Ontological realist epistemological realistOntological relativist epistemological relativistOntological realismOntological relativism (code family)

OR: accounting for individual differences•OR: learning as a process•OR: practice not constrained by curriculum•OR: student centered learning•

Preservice teacher beliefs/practicesShifting beliefsTeacher beliefs impacting practice (code family)

TB: realist•TB: relativist•

Part V

Conclusion

555

17 Personal epistemology in the classroom: what does research and theory tell us and where do we need to go next?

Lisa D. Bendixen University of Nevada, Las Vegas

Florian C. Feucht University of Toledo

Introduction

The impetus of this book is to fill a significant gap in our understanding of the relevance of personal epistemology in preschool through second­ary education. In this concluding chapter we summarize the contribu­tions made by the authors and also consider the broader implications of this work as it pertains to the importance of personal epistemology in the classroom and its role in critical thinking development and the education of our future citizens.

Many of the authors in this book use a systems approach to organ­ize and capture the complexity of personal epistemology. In the sprit of this, we take a systems perspective in this concluding chapter and discuss six themes related to personal epistemology in the classroom. In doing so, we examine and discuss the highlights and themes found in this book, including intriguing empirical and conceptual questions and inroads to future theory and research in the field. The six themes include: (1) conceptual issues; (2) methodological issues; (3) the role of the teacher; (4) educational implications; (5) our educational agenda; and (6) evaluativism (see Figure 17.1).

We have chosen four systems to use as our lens as we discuss the previ­ously listed themes gleaned from this volume. The individual system per­tains to students and teachers themselves and includes, for example, their beliefs, attitudes, abilities, and knowledge strategies. The microsystem includes direct influences on the student such as classroom climate, class­room culture, epistemic climate, and various interactions with students in these contexts (e.g., students–teachers, students–peers, textbooks). The exosystem considers the external, more indirect influences on students. This includes the field and construct of personal epistemology, research/researchers, teacher education/training, curriculum, school context, and

What does research and theory tell us?556

theory associated more specifically with educational psychology. The macrosystem examines aspects of broader contexts that do not influence the student directly but still impact them. Societal and cultural contexts, the general domains of education and psychology, and educational standards (national and international) for students and teachers pertain to this system. As with any systems approach, it is assumed that all of the systems impact one another reciprocally and offer a valuable way to examine these themes regarding personal epistemology in the classroom (see Figure 17.1).

Theme 1: conceptual issues

We concur with Chandler and Proulx (Chapter 7) who state that the field of personal epistemology needs a fair amount of organizing and “cleaning” in its “conceptual house.” All of the authors in this book certainly add clarity to our conceptual understanding of personal epis­temology and they also offer ways in which we can move the field even further in the future.

Individual

Individually, what does the student bring to their own personal epistemology and subsequent learning? This influence was brought out

Evaluativism

Ind. Micro Exo. Macro

Edu. agenda

Edu. implicationsConceptual

Methods

Teacher

Figure 17.1: Six themes related to personal epistemology in the class­room using a systems perspective

Lisa D. Bendixen and Florian C. Feucht 557

in a number of the chapters. A student’s cognitive ability (e.g., formal operational thinking/abstract thinking) as a factor in more advanced epistemological development is discussed (Chandler and Proulx, Chap­ter 7; Schommer­Aikins, Bird, and Bakken, Chapter 2). This does not necessarily mean, however, that elementary school students do not show signs of multiplistic thought, and evidence for this is provided as well (e.g., Greene, Torney­Purta, Azevedo, and Robertson, Chapter 12; Yang and Tsai, Chapter 5). How ability may constrain epistemological development (and vice versa) and how these beliefs manifest them­selves in various age groups has certainly been a part of this volume and should continue to be a focus of future research; in particular, in the use of longitudinal studies.

A student’s own volition or epistemic volition is now being viewed as essential for their own progress as well (e.g., Mason, Chapter 9; Rule and Bendixen, Chapter 4). Giving students the training to consciously act on behalf of their beliefs and development (e.g., what can a student do to resolve their own epistemic questions?) can be a powerful con­tributor to advancement (De Corte, Op ’t Eynde, Depaepe, and Ver­schaffel, Chapter 10; Elby and Hammer, Chapter 13).

The role of the student’s affect in their own epistemological develop­ment continues to gain attention, though much more can be done along these lines in future empirical work. The emotions that coincide with epistemic doubt, for example, can be viewed as a help and/or a hin­drance to progress and this may be due, in part, to how supportive the environment is for students (e.g., Mason, Chapter 9; Rule and Bend­ixen, Chapter 4). In addition, there may be certain times in a student’s development where emotions are more intense and difficult, and this should be recognized and addressed by teachers (e.g., beginning mid­dle school) (Chandler and Proulx, Chapter 7; Perry, 1970).

In general, raising meta­cognitive awareness in terms of the afore­mentioned aspects of epistemic beliefs and its progression is a key aspect of what each student can contribute to their own belief devel­opment and subsequent learning. For example, if a student is aware that, at times, the questions they may have about the historical views of war they are learning about may challenge their current beliefs and possibly make them feel uncomfortable, they might be more likely to communicate this to their teacher and get the support they need to move past it.

Microsystem

It is becoming quite clear that the personal epistemology of students does not operate within a vacuum (Feucht, Chapter 3;

What does research and theory tell us?558

Schommer­Aikins et al., Chapter 2). Specific and intriguing accounts of some of the more direct educational influences on students’ epis­temological beliefs is brought to light in several chapters. In Chapter 10, De Corte et al. provide a very potent account of the classroom cul-ture within the domain of mathematics, including how implicit and explicit classroom norms are communicated and negotiated among students and teachers.

In addition to the student and teacher, other important compo­nents of the epistemic climate or systems within which students oper­ate include family, peers, classroom materials, and curriculum. For instance, the Internet cannot be overlooked as an important and more recent aspect of the classroom context (e.g., Greene et al., Chapter 12; Bromme, Kienhues, and Porsch, Chapter 6). How these influences at the microsystem level interact with one another adds conceptual depth and direction for further study. It is assumed that aspects of the classroom culture interact with one another reciprocally in the form of feedback loops (Rule and Bendixen, Chapter 4; Schommer­Aikins et al., Chapter 2; Yang and Tsai, Chapter 5). For example, Feucht (Chap­ter 3) proposes in his framework that students, teachers, instruction, and educational materials have an epistemic influence on each other in the classroom.

Exosystem

At this level, we focus on conceptual issues that pertain to the field of personal epistemology theory and research. For example, some of the authors provide insight into the question: What is and isn’t epistemologi-cal in personal epistemology?

According to Olafson and Schraw (Chapter 16) and Greene et al. (Chapter 12), epistemological and ontological beliefs need to be consid­ered as related but distinct conceptual entities. Beliefs about learning are also viewed by some authors as intricately tied to epistemological beliefs (DeCorte et al., Chapter 10; Elby and Hammer, Chapter 13; Schommer­Aikins et al., Chapter 2). By and large, personal epistemol­ogy is considered multidimensional in nature (e.g., Wildenger, Hofer, and Burr, Chapter 8), and a number of the authors have begun to tease apart and refine our understanding of these dimensions. For instance, in their model and corresponding measure, Greene et al. (Chapter 12) chose to further distinguish the dimension of justification of knowledge by separating it into justification by authority (i.e., using evidence from experts such as teachers and scientists) and personal justification (i.e., faith in warrants such as personal experience and logic).

Lisa D. Bendixen and Florian C. Feucht 559

In addition, what is “personal” in personal epistemology, according to Bromme et al. (Chapter 6), needs clarification within the field. In discussing “Who’s knowledge is it?” they argue that much of the know­ledge we obtain comes from second­hand knowledge sources (e.g., sci­entists) rather than first­hand or individually constructed knowledge. This is also consistent with Chandler and Proulx’s (Chapter 7) dis­cussion of hand­me­down knowledge or propagated stuff (Hammer and Elby, 2002). Bromme et al. (Chapter 6) also point out that there is more to the dimension structure of knowledge, than just the connected­ness of it. Students, for example, need to also appreciate this dimension in terms of our dependency on those who specialize in certain kinds of knowledge (who knows what?), or have a clear understanding of the division of cognitive labor that exists in our world.

Is there a developmental story? We think so, but we need to be clearer about what we mean regarding how epistemic development unfolds in students. Several of the contributors in this book have pointed out that our current views of epistemological development are too simplified and vague and that there is much of the story left to uncover (e.g., Chandler and Proulx, Chapter 7; Elby and Hammer, Chapter 13). We agree with this sentiment but we also would like to point out that cur­rent works reflected in this volume present a number of clarifications and future directions in terms of understanding the story.

In general, the readers of this book have received an interesting and thorough conceptual and empirical account of personal episte­mology across all of the ages represented in preschool to grade twelve education. The personal epistemology of young knowers (i.e., pre­school/Kindergarten students) is a promising area of new research that carries with it a number of important developmental implica­tions, including the origins of epistemological development. It seems that, as a field, we have oversimplified and underestimated the episte­mological beliefs of children just beginning their formal educational endeavors.

As we see in Chapter 8, Wildenger et al. found definite variability in preschool/Kindergarten students’ beginning personal epistemology, and the idea that they are all absolutists and incapable of understand­ing differing views is not supported in their research (see also Winsor, 2008). More specifically, in investigating the role of age and “theory of mind” (ToM) attainment, they found that three­year­olds and children with low ToM made more relative judgments than did five­year­olds and those with ToM. Further, children became increasingly more abso­lutist as they approached five years of age in tandem with their con­solidation of ToM. The evidence that children, even before the onset

What does research and theory tell us?560

of ToM, are cognitively able to appreciate diversity in beliefs is key to understanding that the beginnings of personal epistemology are unique in and of themselves and not just a byproduct of ToM as was once assumed (Burr and Hofer, 2002; Chandler et al., 2002).

The empirical findings related to elementary students’ personal episte­mology found in various chapters can be characterized as: fairly sophisti­cated within the domain of history (Greene et al., Chapter 12); benefiting significantly from intervention aimed at problem­solving performance and the acquisition of positive beliefs in mathematics (De Corte et al., Chapter 10); able to respond appropriately to epistemic nudging by their instructor (Elby and Hammer, Chapter 13); predominantly absolutist in informal reasoning situations (Yang and Tsai, Chapter 5); directly impacting conceptual change learning (Mason, Chapter 9); and influ­encing their levels of achievement goals (Muis and Foy, Chapter 14).

We are also able to compare and contrast research findings associated with middle school students and other age groups. For instance, mid­dle school students’ epistemological beliefs were shown to be relatively unstable in informal scientific contexts (Yang and Tsai, Chapter 5); to increase meta­cognitive awareness and conceptual change (Mason and Gava, 2007); and serve as significant predictors of grade point average (GPA) and overall goal to learn in low socio­economic status (SES), ethnic minority students (Murphy, Buehl, Zeruth, Edwards, Long, and Monoi, Chapter 11).

Some interesting patterns relating to high school students’ personal epistemology were also found. In reasoning about ill­structured everyday scientific problems, upper secondary students showed signs of evaluativism but displayed mainly multiplistic views. Personal epistemology was also related to these students’ abilities to formu­late their own theories and make inferences based on those theories (Yang and Tsai, Chapter 5). Additionally, high school students’ epi­stemic beliefs and identities related to advanced mathematics were clearly impacted by their classroom cultures (i.e., traditional versus discussion­based instructional approaches) (DeCorte et al., Chapter 10). Finally, beliefs about school work (as opposed to learning) and work­avoidant goals were substantive predictors of achievement in ethnic minority students living in low SES environments (Murphy et al., Chapter 11).

As to the developmental story, in general, evidence is provided to show that epistemological progression does seem to contain relevant patterns but that these patterns are quite complex, diversified, nonlinear, and reliant upon domain and context (e.g., Elby and Hammer, Chapter 13; Greene et al., Chapter 12). For example, Elby and Hammer (Chapter 13)

Lisa D. Bendixen and Florian C. Feucht 561

add epistemological frames to their epistemological resources repertoire and consider these frames to be, at times, fairly stable and belief­like. They also provide many additional insights into students’ personal epistemol­ogy within the context of their own learning in the classroom, and we think this work has great potential for future research.

A number of developmental questions that currently exist in the field are examined as well. The developmental issue of recursion (i.e., why we see the same developmental patterns across various age groups) is one example (Chandler et al., 2002). According to Chandler and Proulx (Chapter 7), if we would focus only on the more relevant epi­stemic aspects of social facts and shared values in our research and forgo the brute facts and personal taste that even young children are able to understand, we might begin to see that there is a more coherent “develop mental story,” and some of our puzzling findings in the field may start to clear up.

Schommer­Aikins et al. (Chapter 2) also provide a slightly different take on recursion through their use of scenarios that include individuals operating within different systems. They propose that there are three time frames for recursion (i.e., revising one’s epistemic beliefs): (1) around five to seven years of age; (2) at the beginning of adolescence; and (3) during young adulthood. This hypothesis would certainly lend itself to addi­tional research.

Other characters in the development story that are discussed include more insight into possible mechanisms of change (Elby and Hammer, Chapter 13) and how regression may be important and prevalent in epistemic development (Rule and Bendixen, Chapter 4; Yang and Tsai, Chapter 5). For example, teachers should be aware that if a student experiences epistemic doubt this does not always guarantee that the student will move forward and quickly resolve it, but that regression back to their former beliefs is likely.

Finally, we see the explicit integration of the dimensional and developmental aspects of personal epistemology as having great promise conceptually, methodologically, and educationally. The seeming “dichotomy” between them that has been portrayed in the past may not be as informative and useful any longer. Hofer (2001) calls for more conceptual work to be done in this regard as well. Greene et al. (Chapter 12), for example, have made significant strides along these lines in their model that incorporates four (developmen­tal) positions (i.e., realism, dogmatism, skepticism, and rational­ism) and three dimensions (i.e., simple/certain knowledge, personal justification, and justification by authority) of personal epistemol­ogy. They “posit that individuals’ beliefs along these dimensions

What does research and theory tell us?562

will vary systematically, indicating their position within the model” (p. 374) (see also Greene et al., 2008). We agree with the authors that important progress can be made when the details of this merger are proposed and examined and it offers much more in the way of conceptual clarity and methodological advancement rather than con­sidering dimensional and developmental theories as opposing views (Bendixen and Haerle, 2005).

Macrosystem

Consideration of the broader societal, cultural, and educational con­texts that students are operating within is definitely apparent in a number of chapters. In our view, these kinds of approaches attempt to capture the complexity of personal epistemology and they offer valu­able insight into the implications of our work ranging from individual student belief change to more general aspects of educational reform. Muis and Foy (Chapter 14) and De Corte et al. (Chapter 10), for instance, propose that national and international standards in math­ematics can and should be examined in terms of their epistemological stances. As these standards influence, among other things, curricu­lum, pedagogy, and student learning in mathematics, they cannot be overlooked.

How does each of the system levels impact one another and vice versa? Aspects of this reciprocity are certainly discussed as a part of each systemic model presented in this volume, but more detail could be provided in the future (e.g., Feucht, Chapter 3). Rule and Bendixen’s (Chapter 4) personal epistemology multiplier (i.e., advances in indi­vidual personal epistemology may trigger a rise in the group average) provides a window into these broader reflexive aspects of a systems approach to understanding personal epistemology.

In addition, it has been astutely pointed out that the value placed on education in general may be different for different groups (e.g., ethnicity, SES) (Murphy et al., Chapter 11) and cultures (Yang and Tsai, Chapter 5) and we must understand these broader differences (and potential similarities) to further our conceptual understanding of personal epistemology.

Theme 2: methodological issues

It is clear that the methods represented in this book are addressing more of the complexity of personal epistemology, especially in how it is

Lisa D. Bendixen and Florian C. Feucht 563

situated within educational contexts (Bendixen and Rule, 2004; Elby and Hammer, Chapter 13; Fives and Buehl, Chapter 15).

Individual

Within the chapters many options and guidance are provided in how to assess individuals’ personal epistemology. As Feucht (Chapter 3) points out, we need to be careful in our methodological choices and not under­estimate younger students’ personal epistemology (like we have done in the past). Wildenger et al. (Chapter 8) found that some aspects of their interviews with vignettes were difficult for their three­year­old partici­pants related to overly complex language that was used and issues asso­ciated with attention span constraints (i.e., recency effect). A number of useful suggestions are made for those who work with younger students including the use of puppets to supplement vignettes, shorter interview length, and more simplified language.

Greene et al. (Chapter 12) utilized cognitive interviewing to clarify their questionnaire items for elementary and middle school students. This technique provided critical information for survey develop­ment and validity purposes. For example, elementary students in their study had a significant amount of difficulty verbalizing their thoughts as they completed the questionnaire, but seemed to have no difficulty with its seven­point Likert scale. In addition, several mis­understandings in terms of the meaning behind certain items were exposed resulting in the rewording and clarification of the items. In addition, Murphy et al. (Chapter 11) found that their adolescent participants had difficulty with the negatively worded Likert items in their measure and call for more research along these lines (see also Murphy et al., 2007).

According to Elby and Hammer (Chapter 13), researchers should not assume stability in students’ epistemological beliefs and our meth­ods need to probe into different contexts and different activities within and across domains to accurately assess how students epistemologically frame a certain learning activity. One way to accomplish this, accord­ing to the authors, would be by utilizing naturalistic case studies and slightly altering experimental conditions in the classroom.

We also agree with Greene et al. (Chapter 12) who point out that many of our studies in the field may be suffering from restriction of range issues (e.g., only using samples of predominantly Caucasian, middle­class college students) and how this can lead to distortions in validity and reliability analyses of measures. Adding more diversity

What does research and theory tell us?564

(e.g., age, ability, ethnicity, etc.) in our samples can help alleviate this problem (Murphy et al., Chapter 11).

Microsystem

When we do go about researching aspects of a classroom culture or epistemic climate, what are some of the methods that lend them­selves to this task? De Corte et al. (Chapter 10) present a plethora of methodological techniques aimed at capturing classroom cultures as they pertain to mathematics. In addition to mathematics, we see their methods as valuable in a number of different educational set­tings and content areas. Several intervention studies are described and most of them start out with a clear goal in mind such as foster­ing more positive mathematics­related beliefs in students. Interven­tion designs are described and include pre/post measures, student and teacher interviews, questionnaires, discourse analyses, and videotap­ing of student problem­solving and classroom discussions. The rigor of their intervention research is also increased by the use of experimental and control group comparisons. Some interventions focus on the use of innovative versus traditional mathematical textbooks or discussion versus traditional instructional approaches. Finally, these authors offer valuable ideas regarding capturing explicit classroom norms using low­inference coding (i.e., concrete, externally observable behaviors) and that of investigating more implicit classroom norms via high­inference coding (i.e., identifying qualities of student­teacher interactions and processes).

Intervention designs, multimethod approaches, and details regarding triangulation are also consistent with other authors (Elby and Hammer, Chapter 13; Mason, Chapter 9). Feucht’s (Chapter 3) model provides not only theoretical backing for examining epistemic climate but guide­lines for methodological pursuits as well. For example, microgenetic research designs could be used not only to assess the product of change but also the actual change processes (e.g., an individual student’s epis­temic development is studied meticulously and constantly over a cer­tain period of time).

Muis and Foy (Chapter 14) focus on aspects of epistemic climate in mathematics at the elementary level by investigating the direct relations between teachers’ epistemological beliefs and that of their students. These direct relations were analyzed using structural equa­tion modeling (specific variables within the models were measured by surveys and a student achievement measure). In researching the links between teachers’ epistemological worldviews and instructional

Lisa D. Bendixen and Florian C. Feucht 565

practices, Olafson and Schraw’s (Chapter 16) four-quadrant scale that employs vignettes, essays, and follow­up interviews allows teach­ers to explain and justify their answers in more realistic educational contexts.

As can be seen, much in the way of researching classroom cultures can be gleaned in the leading­edge methodology represented by the authors in this book.

Exosystem

This section focuses on looking more broadly at some of the methodo­logical issues and innovations in the field of personal epistemology. In general, for more quantitative approaches, Fives and Buehl (Chapter 15) offer ways in which researchers can develop more reliable and valid measures. Their two­pronged approach attempts to take the “complex­ity, specificity, and variations in teachers’ beliefs” and make it more manageable for survey development (p. 503). First, they suggest that researchers “must constrain the domain” in question (in their case, teachers’ beliefs about teaching knowledge) to help ensure that partici­pants (and researchers) are on the same conceptual page (p. 503). Once items are developed that reflect the domain, think­aloud protocols can be utilized to check respondents’ understanding of the domain (see also Greene et al., Chapter 12). Secondly, they suggest that researchers should get an indication of how much respondents value each aspect of the domain in question (e.g., ranking) before proceeding. Taking more deliberate steps such as these in survey development will result in more clarity and validity in our measures.

Other suggestions for future research include a call for methods designed to pull apart and evaluate separate dimensions of epistemo­logical understanding in early epistemic development (Wildenger et al., Chapter 8), and elementary and middle school students (Mason, Chapter 9; Murphy et al., Chapter 11). As has been pointed out, lon­gitudinal studies are still quite rare in our field and are desperately needed (Feucht, Chapter 3). Perhaps with the conceptual and meth­odological advances evident in this volume researchers will be better equipped to take on the important task of longitudinal research.

Finally, several in the field are beginning to talk in terms of devel­oping methodologies that can be used by teachers rather than just for them. For example, in addition to researchers, Elby and Hammer (Chapter 13) describe ways in which teachers themselves can assess how students epistemologically frame a learning activity and try to cap­ture the complexity involved. Consistent with this and the crucial link

What does research and theory tell us?566

between conceptual frameworks and methodology, some of the theoret­ical frameworks in this volume have been developed and/or elaborated upon with the goal of guiding researchers and educators alike (e.g., Feucht, Chapter 3; Greene et al., Chapter 12).

Macrosystem

With our more fluid and complex systems approaches, can we ever cap­ture this complexity in research? In their chapter, Schommer­Aikins et al. (Chapter 2) raise this important question. We agree that this might seem like an overly daunting task, but these authors offer suggestions to help make this task much more reasonable and effective. Their sugges­tions include researchers and practitioners working together to develop conceptual models, and researchers coordinating more in their various methodological strengths and working in teams. Finally, they stress the need for the development of research questions (and their subsequent methodology) that reflect more of a systems approach (e.g., What hap­pens when the epistemological beliefs of parents, teachers, and students differ? How does that impact the student?).

Methodology and research designs that coincide with examining broader influences related to culture, ethnicity, and SES are a must. Examples of progress in these areas are beginning to show themselves in the field including research represented in the present volume (Mur­phy et al., Chapter 11; Yang and Tsai, Chapter 5) and in other recent works (e.g., Khine, 2008). As will be discussed in a later section, we have much to gain in this macrosystem­level inspection of personal epistemology, especially as it pertains to educational reform.

Theme 3: the role of the teacher

Within the complexities of our theory and research on personal epis­temology it is becoming quite apparent that the teacher is vital, an epi­stemic gatekeeper of sorts for students.

Individual

The role of the individual teacher is so very important for students’ developing personal epistemologies and this is acknowledged by a num­ber of chapter authors. “Better education is still, essentially, a matter of teacher characteristics, among them their [epistemic] beliefs” (Mason, Chapter 9, p. 284). This echoes Pajares’s (1992) view that teacher beliefs are more important than anything else (i.e., content knowledge,

Lisa D. Bendixen and Florian C. Feucht 567

skills) in terms of pedagogical choice and subsequent student learn­ing. Olafson and Schraw (Chapter 16) provide evidence that teachers’ beliefs do indeed impact their practice. They also describe a study (Lidar et al., 2005) that found that teachers who were strong relativists had a more student­centered approach to teaching and they also used a greater number of epistemic moves in their science classrooms (i.e., activities that promote generating, constructing, and reconstructing knowledge).

Intriguing insights into the complexity of teachers’ beliefs are also offered. According to Fives and Buehl (Chapter 15), the task of teach­ing is complex so it should be no surprise that teachers’ beliefs about teaching should reflect this as well. They also turn to Schulman’s (1987) categorization scheme of teachers’ knowledge base to illumin­ate some of the intricacies of what teachers know. What do teachers think they need to know? Taking teachers’ perspectives into this as well as researchers’, they propose that teachers’ beliefs about teaching knowledge should be considered a distinct domain and pursue this in their investigations. We concur that this approach provides add­itional depth to our understanding of teachers’ personal epistemology. Similarly, Elby and Hammer (Chapter 13) assert that their framework draws upon and contributes to “teachers’ professional knowledge” (p. 416) (see also Hammer and Elby, 2003) in that it relies heavily on what teachers know about and can recognize in their own students’ learning.

How do teachers’ beliefs about knowledge and knowing develop over time? Are novice teachers more naïve in their personal epistemolo­gies than seasoned teachers? Does the domain in which teachers teach impact their beliefs (e.g., science versus literature)? What are some pos­sible mechanisms responsible for teacher belief change? Questions such as these are examined in addition to the significant need for teachers to reflect and be aware of their own personal epistemologies and how these beliefs may impact their practice (Feucht, Chapter 3; Fives and Buehl, Chapter 15; Olafson and Schraw, Chapter 16).

Microsystem

As is evident in this volume, the teacher is paramount to the epistemic ebb and flow of the classroom climate/culture (e.g., De Corte et al., Chapter 10; Mason, Chapter 9; Rule and Bendixen, Chapter 4). Cer­tainly, we want teachers to understand their own personal epistemol­ogy and that of their students, but it does not stop there. We also want teachers to intervene actively and be a part of epistemic change in their

What does research and theory tell us?568

students in supportive ways. There are many excellent ideas and poign­ant examples of this throughout this volume.

Schommer­Aikins et al. (Chapter 2) walk us through two students’ educational journeys as they pertain to epistemological development and the various contextual influences that are occurring. Several instances are described regarding epistemologically based instruction as well and how teachers can mediate between what is happening in the classroom and students’ beliefs/learning. What is also helpful here is that the authors point out how certain instructional tactics encour­age consideration of specific dimensions of epistemological beliefs. For example, a teacher may encourage students to make connections across concepts, other subjects, and one’s own knowledge (complexity of knowledge) or offer other resources for answers to challenge their belief that information that is portrayed in the media is always correct (source of knowledge).

Other insights are offered relating to how teachers may interject into the epistemic growth of their students, including: setting up incongruities (Rule and Bendixen, Chapter 4; Perry, 1970), asking stu­dents to come up with several ways to solve a problem (Muis and Foy, Chapter 14), establishing epistemologically sound classroom “rules” for problem­solving in mathematics (De Corte et al., Chapter 10), compar­ing differing historical resources (Feucht, Chapter 3), setting up discus­sion/debate regarding conflicting scientific theories (Mason, Chapter 9; Yang and Tsai, Chapter 5), asking even very young students to justify their thinking (Wildenger et al., Chapter 8), and having students iden­tify proper sources/authorities to help understand the division of cogni­tive labor (Bromme et al., Chapter 6).

According to many authors in the book, teachers must have com­passion for their students when they experience the emotional ups and downs of epistemological development (e.g., Fives and Buehl, Chapter 15; Mason, Chapter 9; Perry, 1970). For example, the experience of epi­stemic doubt may have a disruptive effect in students (Rule and Bend­ixen, Chapter 4) or it may spur volition (Mason, Chapter 9). According to Chandler and Proulx (Chapter 7), teachers need to be attuned to this and intervene “where young people’s doubts are felt most acutely” (i.e., learning about the tentative nature of social values and truths) (p. 216). In addition, they also state that researchers should investigate and clarify when this need seems to be the greatest in students and pass this information along to teachers. Similarly, Perry (1978) asked, “Why is it so hard to grow?” and “Why is it even harder to encourage others to grow?” He proposed that with any gain or growth in development there is a loss of innocence for the student and this is not always easy;

Lisa D. Bendixen and Florian C. Feucht 569

teachers need to be aware of this and be compassionate about it for their students.

Exosystem

Teacher education is a significant influence at the exosystem level. Sev­eral chapter authors point to the need for training to include teachers becoming more aware of their own epistemological beliefs (e.g., Mason, Chapter 9) and this needs to be done systematically “as they move their beliefs from tacit to explicit and from transitional to well­developed and enacted” (Fives and Buehl, Chapter 15, pp. 508–9; see also Gill et al., 2004). In addition, more exposure to constructivist teaching prac­tices and ontology training is suggested (Muis and Foy, Chapter 14). Preservice training may be the ideal time for this as teachers’ beliefs may still be forming (Schommer­Aikins et al., Chapter 2). In support of this, Olafson and Schraw (Chapter 16) found that teachers with gradu­ate degrees and more experienced teachers tended to be multiplistic in their beliefs (rather than absolutist).

An additional form of awareness includes teachers understanding that they are often viewed as an epistemological authority by their stu­dents (Greene et al., Chapter 12) and that just by being a teacher they are viewed as having “a special corner on the truth” (Chandler and Proulx, Chapter 7, p. 208). As we will discuss later, teachers can and should be the authority sometimes, but they also must be cognizant of this possibly exaggerated belief and the responsibilities that go along with it.

Mason (Chapter 9) and Feucht (Chapter 3) also propose that an important part of education for teachers should include an under­standing of the epistemological basis of the subjects/content/discipline they teach. Only when a teacher “knows the questions a discipline deals with are they able to know” how that discipline considers, for example, certainty within that area (Bromme et al., 2008). For exam­ple, how is mathematical knowledge typically viewed in educational settings? How do these broader epistemic views of mathematics (or any other content area), in turn, affect how teachers are expected to teach and assess students (De Corte et al., Chapter 10; Muis and Foy, Chapter 14)?

Elby and Hammer (Chapter 13) see their framework as useful for teacher training because it is based within the classroom context and provides clear­cut guidance for experienced teachers to be able to rec­ognize, and what novice teachers can learn to recognize, in their stu­dents’ epistemological framing (e.g., what sort of activity is this?). Fives

What does research and theory tell us?570

and Buehl (Chapter 15) also point out that there may be the danger of teachers merely developing the rhetoric from their teacher education programs and some teachers may be resistant to change due to certain epistemological beliefs (Mason, Chapter 9; Olafson and Schraw, Chap­ter 16; Patrick and Pintrich, 2001).

Macrosystem

The everyday realities of teaching are brought to light by Olafson and Schraw (Chapter 16) and it seems clear that teachers are operating within a complex web of influences that go beyond their individual beliefs about knowledge and knowing and can be viewed from both the exosystem and the macrosystem level. The teachers in their research describe influences such as administration, overly structured programs, accountability for practice, funding, testing, time, and the general school context/climate. The picture they paint includes teachers feel­ing that they are being stripped of their instructional decision­making power due to district­mandated curriculum driving their practice and the pressures of standardized testing. Interestingly, a small number of teachers in their study felt that the structure that was provided was actually helpful in their teaching rather than restrictive. Interestingly, this may be due, in part, to these teachers’ beliefs about knowledge and knowing.

As can be seen, the broader sources of influence that are embed­ded within the expectations for teachers and their teaching need to be examined epistemologically. For example, the National Council for Accreditation of Teacher Education (NCATE) includes standards for teachers and the National Council of Teachers of Mathematics (NCTM) has called for teachers to implement constructivist teach­ing techniques in mathematics to increase student performance (Muis and Foy, Chapter 14). What these influences mean for teachers’ per­sonal epistemologies and the learning of their students must be exam­ined further by the field. Teacher training, for instance, could help teachers operate more successfully in a standards­based educational world.

Finally, additional and very potent questions to scrutinize include “What are teachers’ beliefs about the role of school in society?” and, more ontologically, “What does it mean to be a good teacher?” ( Olafson and Schraw, Chapter 16). Answers to these questions should be investi­gated and the findings could be very telling in terms of society’s expec­tations of teachers and teachers’ expectations of themselves.

Lisa D. Bendixen and Florian C. Feucht 571

Theme 4: educational implications

Educational implications of personal epistemology theory and research that are outside of the direct consideration of the teacher will be dis­cussed in this section.

Individual

What kinds of activities do students need in the classroom? Many suggestions were provided by the authors in this book pertaining to this question. According to Yang and Tsai (Chapter 5), more reason­ing experiences with ill­defined problems in informal contexts should be offered to students. Real science experiments – ones that do not always turn out the same way with one solution – could also be a part of student learning as opposed to low­level experiments (Bromme et al., Chapter 6; Mason, Chapter 9).

Many examples and suggestions exist in terms of activities that encour­age more student involvement/volition, reflection, and epistemological discussion. For example, active student engagement could be encour­aged by argumentation and debate (Mason, Chapter 9) and more reflec­tion could be achieved through journaling and discussion (Rule and Bendixen, Chapter 4). Elby and Hammer (Chapter 13) provide explicit guidance in activities that encourage students to “practice” thinking more epistemologically about what it is they are learning. This could be done through student framing as they get used to asking questions such as “What sort of activity is this?” and “What is going on with respect to knowledge?” (pp. 413 and 414). In addition, Schommer­Aikins et al. (Chapter 2) give a number of suggestions for student activities that are more developmentally appropriate at the elementary, middle, and high school level. We also see that activities such as these are valuable not only for learning but also for convincing students that they can take volitional control of their beliefs.

Students partaking in various kinds of evaluation during their learning is also brought out as a key. For instance, evaluation of theory, evidence, and second­hand sources such as those found on the Internet allow stu­dents to grapple more with the relative nature of knowledge (Bromme et al., Chapter 6; Mason, Chapter 9; Yang and Tsai, Chapter 5).

Finally, a major educational implication that we feel stems from the works represented in this volume is the intentional role of the students in their own learning and epistemic belief change (e.g., Mason, Chapter 9; Rule and Bendixen, Chapter 4). As Kuhn (2005) also succinctly points out, “only the holder of the belief can change it” (p. 16).

What does research and theory tell us?572

Microsystem

The epistemic representations of classroom materials is another addi­tional and crucial layer to understanding the epistemic climate of class­rooms (Feucht, Chapter 3). We see the use of textbooks, for example, as a valuable epistemic opportunity in the classroom. Even traditional textbooks that do not explicitly provide epistemological questions or ones that offer more absolutist views of a content area can be read, discussed, and/or challenged by teachers and students (Feucht, 2008). In other words, teachers can provide “supplements” (e.g., news media, expert opinion) to the traditional classroom text and encourage/model more epistemic discussion regarding their content (e.g., “How does the text book’s description of Pluto as a planet differ from the more recent scientific reports?”). Similarly, as Mason (Chapter 9) points out, sci­ence textbooks usually omit the controversial nature of most scientific progress and only present it in a non­problematic form. Again, even if this is the case, a teacher could use this more “certain” and “simple” textbook portrayal of science knowledge as an example for students to contemplate and discuss.

Additional suggestions pertaining to textbooks and other classroom materials include analyzing how knowledge is represented in texts dif­ferently in different domains (e.g., science versus psychology) (Bromme et al., Chapter 6; De Corte, Chapter 10) and the call for providing more critical reading opportunities for students (Bromme et al., Chapter 6).

In their case study, Elby and Hammer (Chapter 13) provide a detailed window into the classroom discourse going on between students, the teacher, and the use of a worksheet. With the teacher’s guidance (i.e., epistemic nudging), the students over a period of time progressed from using the worksheet as a source for the answer to using it as a tool in their own meaning making. This example also speaks to the value of using classroom materials to enhance personal epistemology and learning.

The Internet has also surfaced as a potential and valuable epistemic tool for instruction (Yang and Tsai, Chapter 5). How students can practice using and evaluating knowledge provided on the Internet and the finding that it is viewed by some students as a more reliable source of knowledge than the teacher and/or textbooks, are certainly topics worthy of further research (Bromme et al., Chapter 6; Greene et al., Chapter 12; Mason, Chapter 9).

What role does classroom assessment play in the personal epistem­ology of students? What kind of epistemic messages are embedded in assessments? It is often suggested that students be given assessments that require more critical thinking and other higher forms of cognition

Lisa D. Bendixen and Florian C. Feucht 573

to enhance their learning (Anderson and Krathwohl, 2001; Bloom, 1984; Stiggins, 2004). We also see that an over­reliance on low­level assessments can be detrimental to students’ views of knowledge and knowing as well (Haerle and Bendixen, 2008). In mathematics, for example, students are rarely given assessments that go beyond finding the one correct answer (De Corte et al., Chapter 10; Muis and Foy, Chapter 14). In any content area, if all that is required of students is factual and basic comprehension, their views that knowledge is mostly simple, certain, and not in need of justification is certainly understand­able. Assessment techniques that allow for more “productive” personal epistemologies in students cannot be overlooked.

Finally, other individuals operating within the microsystem level include peers and parents. The ways in which each of these groups can enhance or hinder a students’ personal epistemology is important for researchers and educators to consider. For example, peer criticism can be an effective part of the epistemic climate of the classroom (Rule and Bendixen, Chapter 4; Schommer­Aikins et al., Chapter 2).

Exosystem

So, it would appear, everything – or at least everything potentially educational – comes down to so­called “social” or “institutional” or “value­impregnated” facts. Here, there is good reason to suspect, education has the possibility of getting a real toehold. (Chandler and Proulx, Chapter 7, p. 215)

It should be no surprise that the brute facts Chandler and his col­leagues speak of are not lacking in coverage in today’s classrooms and the standardized testing movement.

As was mentioned earlier, having students be exposed to social facts, or knowledge that has been evaluated and agreed upon by others as true, brings with it the potential for higher levels of thinking and more advanced epistemological viewpoints. In a sense, this is not a new mes­sage. There are certainly parallels among this message and what has been advocated for in the educational arena including pushes for more critical thinking (Barnes, 1970; Paul et al., 1990), reflective judgment (King and Kitchener, 1994), thinking curriculum (Resnick and Klopfer, 1989), and higher levels in Bloom’s Taxonomy (Anderson and Krath­wohl, 2001). We think it is imperative to closely examine these parallels between what we are advocating for in personal epistemology research and other movements already going on in education. The more we can band together, so to speak, the better our chances for having our field be taken seriously by those operating within education circles.

What does research and theory tell us?574

The context of school is an important facet of the exosystem and we get a glimpse into the contexts of preschool, elementary, middle, and high school and how these may impact students’ personal epistemology. It would appear that the context of preschool is in flux with pressures to “push down” the more academically oriented Kindergarten curriculum which would also increase the reliance on more traditional, less develop­mentally appropriate instructional practices. We agree with Wildenger et al. (Chapter 8), who point out that these changes in preschool education may induce for students a climate of more absolute/certain knowledge and skills that must be covered to meet National Association for the Education of Young Children (NAEYC) standards and the No Child Left Behind (NCLB) initiative. The epistemic climate of elementary schools is where basic knowledge (e.g., mathematics, reading, writing) is taught with an emphasis on the individual student and more student–teacher interaction (Feucht and Bendixen, in press; Haerle and Bendixen, 2008).

In comparing the contexts of middle school and high school, Murphy et al. (Chapter 11) describe middle school as more collaborative (e.g., students and teachers working in pods) and high schools as containing larger classes with more of a competitive academic focus. In addition, we must investigate the transitions that take place for students within these contexts, such as elementary to middle school, and middle school to high school. These may be delicate epistemological and emotional times for students (Mason, Chapter 9; Yang and Tsai, Chapter 5). In general, how these school contexts may influence, and be influenced by, students’ per­sonal epistemology is an important direction for future research.

Finally, a broader understanding of the subjects they are learn­ing about has a great deal of epistemological potential for students. Bromme et al. (Chapter 6) describe this as a “comparative approach to study subjects” and this can also be accomplished in teaching across the curriculum or taking a more interdisciplinary/integrative approach to curriculum that attempts to look at disciplines and not just topics (p. 186; see also Stevens et al., 2005).

Macrosystem

Personal epistemology plays a part in the cultural and societal aspects of education in terms of what is valued. In essence, effective education entails not just skills but values (Kuhn, 2005). For example, do students believe that what they do in school is relevant to their current and future lives (Murphy et al., Chapter 11)? This can be considered more at the societal level as well. For example, what kind of knowledge is valued? Are there important ethnic and cross­cultural similarities and differences

Lisa D. Bendixen and Florian C. Feucht 575

along these lines as well (Feucht and Bendixen, in press; Haerle and Ben­dixen, 2008; Hofer, 2008)? We must examine these broader influences more closely in our future work.

In general, it seems that education has certain responsibilities to students. For example, Perry (1970) asserts that education should equip students for the relativism that exists in society. Bromme et al. (Chapter 6) and Kuhn (2005) echo this in that we must educate our students to be able to navigate our rapidly changing, complex, and technologically advanced world.

The educational implications that stem from standardized testing/high­stakes testing have been brought up a number of times through­out the chapters (e.g., Muis and Foy, Chapter 14; Olafson and Schraw, Chapter 16). In their current form, standardized testing, or the broader “assessment culture” that we are operating in, does not seem to reflect the messages found in this book about encouraging higher levels of thinking and personal epistemology (Haerle and Bendixen, 2008, p. 166). As Kuhn (2005) firmly states, “Ironically, the current preoccu­pation with standardized test scores, motivated by a concern over the quality of the educational system, is probably the single greatest factor that will constrain its development over the next decade.” (p. 10). We believe that, as a field, we can make significant contributions in over­coming this damaging fixation.

Theme 5: what is our educational agenda?

As we continue to make strides in our understanding of personal epis­temology and its educational implications, we think, as a field, the time is right to thoroughly examine and communicate our educational goals to each other and to the broader educational community. As research­ers in the field of personal epistemology, for example, what is our edu­cational agenda? What are some of the educational goals we hope our work will accomplish? How do we communicate this to others? What should teachers strive for as they use what is known about personal epistemology in their classrooms? We propose that, in general, our edu­cational agenda should explicitly include, as a main goal, the epistemo­logical advancement of students and teachers.

Individual

What does epistemological advancement look like in students? If epistemo­logical advancement is a goal, then we must, of course, be clear on what it is we are striving for. There are a number of examples of epistemological

What does research and theory tell us?576

advancement in the chapters, including students becoming able to con­sider theory and evidence in informal reasoning tasks (Yang and Tsai, Chapter 5), making informed choices when confronted with competing theories (Mason, Chapter 9), and generating several ways of solving a math problem and then deciding which one is best (Rule and Bendixen, Chapter 4). In general, fostering more epistemic sophistication in stu­dents is essential because it relates to better learning outcomes for them (De Corte et al., Chapter 10; Muis and Foy, Chapter 14).

Are students aware of the educational and epistemological goals that others have for them? We believe that, in general, this is not the case. If there is better communication regarding goals, students will be more likely to be engaged in their own learning and beliefs. As De Corte et al. (Chapter 10) and Feucht (Chapter 3) point out, goals and norms within the classroom culture can be communicated both explicitly (e.g., dis­cussion/debate about goals) and implicitly (e.g., epistemic messages found in instructional practices) for students. Ideally, students would become involved in the goal­setting for the group and would also be active in the goals set for themselves.

Microsystem

The classroom culture or epistemic climate must also be conducive to the goals that support epistemic advancement. If a teacher has as one of their goals the epistemic advancement of their students, what would their classroom look like? In reading through the chapters in this volume, a glimpse at what a classroom such as this may look like is provided. Rule and Bendixen (Chapter 4) and Schommer­Aikins et al. (Chapter 2) offer descriptions of fictitious teachers who seem to set up their classrooms with epistemological goals in mind. Although she did not explicitly discuss “epistemological advancement,” Ms. Phelan, the actual teacher described in Elby and Hammer’s (Chapter 13) research, certainly is an excellent example of how she helped her students to think about (frame) the knowledge they were working with.

As has been pointed out, certain subject/domain/discipline areas may have different epistemological goals (science versus mathematics) and teachers (and students) should understand these as part of their expert­ise in developing their classroom cultures (Bromme et al., Chapter 6; Feucht, Chapter 3; Muis and Foy, Chapter 14).

Exosystem

Muis and Foy (Chapter 14) point to the view of Gunzenhauser (2003), who states that there is a “lack of reflective and engaged dialogue

Lisa D. Bendixen and Florian C. Feucht 577

among educators and school communities about their goals and prac­tices” (p. 31). We agree and would add educational psychologists to this list as well; we need to be better at communicating with the educational community.

Hopefully, as a field, we are not falling prey to a “sea of relativity” (Chandler et al., 1990) when it comes to thinking about and express­ing our educational goals to others. Are some epistemological beliefs “better” than others in terms of fostering learning? In our view, there is enough empirical evidence in the field, some of which is represented in the leading­edge research contained within this book, to answer yes to this important question. This book is testament to the fact that we are indeed getting better at discussing our agendas and taking more of a stand on the goals of epistemological advancement.

Another advantage to having and stating our goals more explicitly would be additional clarity regarding the construct of personal epis­temology and more guidance for future research in the field. If we truly know what we are striving for, then ways to measure students’ epis­temological beliefs or direction that we would give in terms of teacher education would be better served.

The epistemological underpinnings of curriculum (Feucht, Chapter 3) also operate at the exosystem level. As has been stated, the epistemic messages behind curriculum must be investigated and understood as well as its potential power. For example, Olafson and Schraw (Chap­ter 16) provide evidence that a number of teachers view curriculum as driving their teaching (and, therefore, their students’ personal epis­temology and learning). For us, it seems fair to say that most of the curriculum in schools today is not significantly in line with the goals of epistemological advancement (Haerle and Bendixen, 2008; Kuhn, 2005). Therefore, it might be quite beneficial for those of us in the field to work with individuals and groups involved in curricular develop­ment. In other words, if curriculum is so powerful, let’s be a part of changing it for the better.

Macrosystem

Why is the goal of epistemological advancement important at the soci­etal and cultural level? Many of the works in the chapters speak to the importance of advancing students’ epistemological beliefs because they will eventually be contributing members of society (Feucht, Chapter 3; Rule and Bendixen, Chapter 4; Schommer­Aikins et al., Chapter 2). As members of a knowledge­based society, for example, we need the skills and beliefs to make informed choices about the theory and evidence

What does research and theory tell us?578

we are exposed to everyday (Bromme et al., Chapter 6; Yang and Tsai, Chapter 5).

Is the goal of epistemological advancement the same across cultures? To help interpret their research findings, Yang and Tsai (Chapter 5), for example, state that there is more pressure to prepare for high school entrance examinations at the eighth­grade level as compared to other grade levels in Taiwan (and other Asian countries). They see this as an important influence on the more teacher­centered instruction that may have been occurring and, subsequently, the less sophisticated epis­temological beliefs demonstrated in these middle school students. This aspect of culture help us to understand their findings specifically, and the impact of culture in general.

According to Hofer (2008), the consideration of culture helps to broaden our understanding of the construct of personal epistemol­ogy. As many of the models in our field are grounded in Western views of education, we must be careful not to make too many assumptions regarding the goals of epistemological advancement without investigat­ing more the influence of culture on views of knowledge and knowing (Feucht and Bendixen, in press).

In terms of communicating our educational agenda in general, sev­eral of the chapter authors speak of their work in terms of its implicit links to broader education goals such as rational reasoning, cognitive advancement, and critical thinking. We think this is a step toward pro­gress in terms of better communicating epistemological advancement as an educational goal as well. The more we can exchange ideas about how personal epistemology could influence policy­makers, for example, the more possible actual change could be (Haerle and Bendixen, 2008). For instance, how often do we, as a field, speak of educational reform? Do we see the goal of epistemological advancement as a viable way to impact reform? The recent work of Kuhn (2005) is an excellent example of how what we do in the field of personal epistemology can translate into edu­cational reform. We believe that the pursuit of our part in educational change should be an important aspect of our identity as a field.

Theme 6: evaluativism

If our goal is epistemic advancement, do we have a clear understanding of the upper echelons of epistemological development and what students and teachers should be striving for? What do we mean by advanced/sophisticated/availing epistemological beliefs? In our view, we need to be much more transparent on this particular aspect of personal epis­temology before our goals can truly be attained. The conceptual and

Lisa D. Bendixen and Florian C. Feucht 579

empirical works in the chapters in this book demonstrate the progress we are making along these lines, but we believe more work is required.

We are choosing the concept of evaluativism put forth by Kuhn and her colleagues (e.g., Kuhn and Weinstock, 2002) to represent more advanced epistemological beliefs/thinking as do several of the chap­ter authors (e.g., Bromme et al., Chapter 6; Feucht, Chapter 3; Rule and Bendixen, Chapter 4; Yang and Tsai, Chapter 5). To briefly review Kuhn and colleagues’ framework, the patterns of epistemological develop ment have been described as occurring in three distinct forms of thinking about the nature of knowledge and the process of knowing. In the first, absolutist form (i.e., objective), views about knowledge are very simple and dichotomous; truth is judged based on an objective, external reality. The multiplistic nature of knowledge (i.e., subjective) is the focus in the second form of thinking where each claim is consid­ered equally legitimate and, therefore, cannot be judged beyond mere opinion. The third, evaluativistic form, of epistemic thinking integrates the objective and subjective nature of knowledge and considers how differing viewpoints can be judged based on established criteria (Kuhn and Weinstock, 2002).

More specifically, then, the educational agenda we are proposing includes epistemological advancement toward evaluativistic thinking in students and teachers. We agree with a number of the authors who assert that we are too vague and simple in our conceptual understandings, especially as they pertain to our understanding of evaluativistic think­ing (Chandler and Proulx, Chapter 7; Elby and Hammer, Chapter 13). The following section attempts to raise pertinent questions and pro­vide discussion to aid in further clarifying the construct, especially as it pertains to evaluativism.

Individual

There is a strong need to investigate students who are evaluativistic thinkers (Yang and Tsai, Chapter 5). What does evaluativism look like in students? Throughout this book these kinds of beliefs have been given different monikers including availing beliefs (De Corte et al., Chapter 10; Muis, 2004), a mature understanding of social facts and values (Chandler and Proulx, Chapter 7), and the coordin­ation of theory and evidence in informal reasoning (Yang and Tsai, Chapter 5).

Interestingly, several studies in this volume report that some form of evaluativism was displayed by elementary through secondary students. For example, aspects of evaluativism in elementary students have been

What does research and theory tell us?580

found although it does not seem to be overly prevalent (Yang and Tsai, Chapter 5; Greene et al., Chapter 12; Haerle, 2006; Mason, Chapter 9). It could be that due to developmental differences in cognition, for example, there may be different versions of evaluativism across the ages, but we certainly can no longer assume that all children prior to late adolescence are completely absolutist in their thinking. How evalu­ativism may be similar and different across various age groups is an intriguing and crucial question for future research. Or, could it be that if we only examine more “pure” epistemic beliefs (i.e., social facts and shared values) in our research, these evaluativistic aspects of children’s thinking would disappear and be seen only in older adolescents and adults (Chandler and Proulx, Chapter 7)?

Investigating the evaluativtisic thinking of teachers is also an area we should understand more. In their research on practicing teachers, Olaf­son and Schraw (Chapter 16) assert that evaluativistic thinking could possibly be represented somewhere in the middle of their four­quadrant scale and not at its extreme endpoints (strong realism or strong relativ­ism). As with research on students, we need to make additional advance­ments in how we measure teachers’ personal epistemology, especially as it pertains to evaluativistic views.

Microsystem

Classroom cultures and epistemic climates conducive to evaluativism should be further explored and understood by the field. There are cer­tainly pockets of these kinds of climates everywhere, but a more sys­tematic approach to evaluativism in the classroom needs to be further explored, understood, and advocated for (Feucht, Chapter 3).

If our goals toward evaluativism for students become clearer, the role of the teacher in fostering epistemic doubt and subsequent change becomes more explicit and realistic as well (Rule and Bendixen, Chap­ter 4). For instance, if understanding the nuances of evaluativistic thinking in students becomes a part of teachers’ expertise and goals, their chances of successful intervention become that much more likely. This was certainly evident in the intervention research reported on by De Corte et al. (Chapter 10).

According to Kuhn (2005), the transition from absolutism to multi­plism may require less tending by teachers. The transition from multi­plism to evaluativism, however, is where the challenge lies. This seems to coincide with Perry’s (1970) and Chandler and Proulx’s (Chapter 7) view that certain times in epistemological development are more diffi­cult than others, and educators must be cognizant of this.

Lisa D. Bendixen and Florian C. Feucht 581

We certainly see promise in the links between evaluativistic thinking and constructivism (i.e., recognition of the tentative/complex nature of knowledge and the active construction of knowledge) (Muis and Foy, Chapter 14). Such teacher beliefs and instructional strategies seem to be consistent with evaluativism, but this association requires further investigation.

Exosystem

As a field, we are still quite ambiguous about the conceptual under­standing of the evaluativist orientation (Wong et al., 2008).

As has been discussed, evaluativism consists of the integration of objectivity and subjectivity in knowledge and knowing (Kuhn and Weinstock, 2002). What does this coordination mean? How do these two aspects become coordinated? For instance, Wildenger et al. (Chap­ter 8) state that “the evaluativist achieves a mature balance” between objectivism and subjectivism (p. 222). In addition, Mason (Chapter 9) refers to this coordination as “neither aspect dominating the other.” Chandler and Proulx (Chapter 7) describe it as a “serviceable mid­dle ground” between the two (p. 213). We argue that this coordination needs much more conceptual and empirical clarity.

We propose that evaluativism is something qualitatively different than just the mere combination of objective and subjective beliefs about knowledge and, therefore, it cannot be just a simple 50–50 balance between the two at all times. It could be that, depending on the context one is operating within, objective and subjective beliefs are not always in balance but that they shift in importance but still work in tandem, so to speak. In essence, this coordination does not look or behave in the same way on all occasions. Sometimes one might rely more heav­ily on objective aspects of knowledge, while at other times they may be more equal. Or, a situation may require more subjective aspects of knowledge to dominate. For example, if a student is learning about the interpretation of poetry in high school they might rely more heavily on the subjective nature of this endeavor. This is not to say that the inter­pretation of poetry is completely subjective and that there is no way to determine the quality of one poem over the next. On the contrary, a student can acquire more objective skills to interpret quality in poetry, but the subjective aspects of poetry knowledge may still play a bigger role (Zemp, 2009).

This more differentiated view of evaluativism may also help with some of the perceptive criticisms raised by the authors in this book. We concur with those who state that we need to avoid the simple

What does research and theory tell us?582

characterization of students’ epistemological sophistication. What is sophisticated epistemology they ask? Always viewing knowledge as ten­tative (i.e., a subjective view of knowledge), is not always the most help­ful and may be even counterproductive for students (e.g., Chandler and Proulx, Chapter 7; Elby and Hammer, Chapter 13; Schommer­Aikins et al., Chapter 2). For example, it is “just not sophisticated to consider it tentative that the Earth is round” (Hammer and Elby, 2002, p. 186). This is, of course, where the coordination of subjective and objective beliefs of evaluativism comes in. As we pointed out, within the con­text of scientific discovery, for instance, it makes sense to rely on the expertise of scientists a bit more heavily (i.e., objective), but this is not a total and blind deference to authority (as in absolutism), but rather an informed choice.

Bromme et al. (Chapter 6) also point this out, as they discuss our need as students and adult members of society to rely on the know­ledge of experts (i.e., second­hand knowledge) most of our lives. It does not seem plausible to constantly construct knowledge completely on our own or to doubt well­established knowledge, but instead to develop skills to make informed choices regarding “who to believe” when it comes up.

In terms of research, we seem to be fairly clear in our understanding of what absolutist and multiplist beliefs look like in students and teachers. We also seem to do a fairly good job of measuring them in the field (e.g., Likert scales have used the absolutism (objective) – relativism (subjective) continuum for a number of years). In terms of what is “beyond” multiplism, however, things get a bit murky in our conceptual understanding and measurement. It seems that Likert scale problems abound when it comes to clearly measuring evaluativistic beliefs. For example, where is evaluativism represented on a typical Likert scale? In the middle, or somewhere beyond the continuum of relativism? (Bendixen, 2004; Hofer, 2004). A more detailed and accurate conceptual understanding of evaluativism can only help our pursuit of instruments that attempt to capture personal epistemology.

In terms of the educational implications of a more detailed view of eval­uativism, there are many. According to Elby and Hammer (Chapter 13), there are times when a transmission approach to teaching can be a pro­ductive part of constructivism. Right now, the constructivist view of teaching is too simple to achieve the goals of evaluativism. Teachers, for example, are advised “not to tell” students the answers to problems in hopes of encouraging students to construct their own knowledge (p. 425). Teachers should rather rely on their expertise and knowledge of

Lisa D. Bendixen and Florian C. Feucht 583

students to make this judgment about when and where it is appropriate to be the authority. Similarly, Perry (1970) asks, when is it okay to cor­rect students and when should teachers be more of a guide? This, again, would depend on the context of the learning situation and the content area in question. In our view, these extremely key educational ques­tions all harken back to clarity in what evaluativism looks like in the classroom.

Macrosystem

As we have discussed elsewhere, evaluativism is a very crucial part of broader educational goals and implications for society. “We argue that evaluativistic thinking is most evocative for, and expressive of, the idea of democracy and what makes a well­informed citizen in West­ern cultures” (Haerle and Bendixen, 2008, p. 167). How this applies cross­culturally is an additional area in need of further understanding as globalization becomes more and more prevalent and reliant upon education (Khine, 2008).

This view of evaluativism is also in line with Bromme et al.’s (Chapter 6) view of the goals of general education (i.e., basic education for all mem­bers of society). In this sense, society is requiring that its citizens be equipped with evaluativistic skills. If we do need to depend heavily on experts in various areas due to the division of cognitive labor and spe­cialization in society, we must make sure our students and future citi­zens have the skills to navigate through information technology, work with authority and expertise, and work with it well.

Kuhn (2005) also discusses educating for citizenship, and that it can be a complex and overly difficult task, especially if the goal is one kind of citizenship for all. Rather, a more effective way to think of this would be “to prepare youths to engage in effective debate of the issues that arise in a democratic society that coexists with a diversity of other soci­eties in a complex world” (p. 11).

We do agree that educating for citizenship is not an easy task but it is one that can be accomplished. Our goal for this book was to expand and possibly reshape the boundaries of personal epistemology theory and research. In addition, we challenged ourselves and our colleagues to place our work in personal epistemology more within the educational arena and we believe these goals have been accomplished. Our work, however, is not over; it is just beginning. If educating our future citizens is also a mission we are willing to take on as a field, then even more clar­ity must be gained in terms of the promise of personal epistemology and evaluativism in the classroom.

What does research and theory tell us?584

R EF ER ENCES

Anderson, L. and Krathwohl, D. (2001). A taxonomy for learning, teaching and assessing: A revision of Bloom’s taxonomy of educational objectives. New York: Longman.

Barnes, D. L. (1970). Identifying and using critical thinking skills in the elementary classroom. ERIC Document Reproduction Service No. EJ034804.

Bendixen, L. D. (2004). Discussant. In M. Schommer­Aikins (Chair), Sym­posium conducted at the annual meeting of the American Educational Research Association, San Diego, CA.

Bendixen, L. D. and Haerle, F. (2005). An interactive measure of children’s per-sonal epistemology. Paper presented at the 11th conference of the European Association for Research on Learning and Instruction. Nicosia, Cyprus.

Bendixen, L. D. and Rule, D. C. (2004). An integrative approach to personal epistemology: A guiding model. Educational Psychologist, 39(1), 69–80.

Bloom, B. (1984). Taxonomy of educational objectives. Boston, MA: Allyn and Bacon.

Bromme, R., Kienhues, D., and Stahl, E. (2008). Knowledge and epistemo­logical beliefs: An intimate but complicated relationship. In M. S. Khine (Ed.), Knowing, knowledge and beliefs (423–41). New York: Springer.

Burr, J. E. and Hofer, B. K. (2002). Personal epistemology and theory of mind: Deciphering young children’s beliefs about knowledge and know­ing. New Ideas in Psychology, 20, 199–224.

Chandler, M., Boyes, M., and Ball, L. (1990). Relativism and stations of epis­temic doubt. Journal of Experimental Child Psychology, 50, 370–95.

Chandler, M. J., Hallett, D., and Sokol, B. W. (2002). Competing claims about competing knowledge claims. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (154–68). Mahwah, NJ: Lawrence Erlbaum Associates.

Feucht, F. C. (2008). The nature of epistemic climates in elementary classrooms. Unpublished doctoral dissertation. University of Nevada, Las Vegas.

Feucht, F. C. and Bendixen, L. D. (in press). Exploring similarities and differ­ences in personal epistemologies of US and German elementary school teachers. Cognition and Instruction.

Gill, M. G., Ashton, P. T., and Algina, J. (2004). Changing preservice teachers’ epistemological beliefs about teaching and learning in mathematics: An intervention study. Contemporary Educational Psychology, 29, 164–85.

Greene, J. A., Azevedo, R., and Torney­Purta, J. (2008). Modeling epistemic and ontological cognition: Philosophical perspectives and methodological directions. Educational Psychologist, 43(3), 142–60.

Gunzenhauser, M. G. (2003). High stakes testing and the default philosophy of education. Theory into Practice, 42, 51–8.

Haerle, F. C. (2006). Personal epistemologies of fourth-graders: Their beliefs about knowledge and knowing. Oldenburg, Germany: Didaktisches Zentrum.

Haerle, F. C. and Bendixen, L. D. (2008). Personal epistemology in elementary classrooms: A conceptual comparison of Germany and the United States and a guide for future cross­cultural research. In M. S. Khine (Ed.), Knowing, knowledge and beliefs: Epistemological studies across diverse cultures (151–76). New York: Springer.

Lisa D. Bendixen and Florian C. Feucht 585

Hammer, D. H. and Elby, A. (2002). On the form of personal epistemology. In B. K. Hofer and P. R. Pintrich (Eds.), Personal epistemology: The psychology of beliefs about knowledge and knowing (169–90). Mahwah, NJ: Lawrence Erlbaum Associates.

(2003). Tapping epistemological resources for learning physics. Journal of the Learning Sciences, 12(1), 53–90.

(2001). Personal epistemology research: Implications for learning and teach­ing. Educational Psychology Review, 13, 353–83.

Hofer, B. K. (2004). Epistemological understanding as a metacognitive pro­cess: Thinking aloud during online searching. Educational Psychologist, 39, 43–55.

(2008). Personal epistemology and culture. In M. S. Khine (Ed.). Knowing, knowledge and beliefs: Epistemological studies across diverse cultures (3–22). New York: Springer.

Khine, M. S. (2008). Knowing, knowledge and beliefs: Epistemological studies across diverse cultures. New York: Springer.

King, P. and Kitchener, K. S. (1994). Developing reflective judgment: Under-standing and promoting intellectual growth and critical thinking in adolescents and adults. San Francisco: Jossey­Bass.

Kuhn, D. (2005). Education for thinking. Cambridge, MA: Harvard University Press.

Kuhn, D. and Weinstock, M. (2002). What is epistemological thinking and why does it matter? In B. K. Hofer and P. R. Pintrich (Eds.), Personal epis-temology: The psychology of beliefs about knowledge and knowing (121–44). Mahwah, NJ: Lawrence Erlbaum Associates.

Lidar, M., Lundqvist, E., and Ostman, L. (2005). The interplay between teachers’ epistemological moves and students’ practical epistemology. Sci-ence Education, 89, 149–63.

Mason, L. and Gava, M. (2007). Effects of epistemological beliefs and learn­ing text structure on conceptual change. In S. Vosniadou et al. (Eds.), Reframing the conceptual change approach in learning and instruction (165–96). Oxford, UK: Elsevier.

Muis, K. R. (2004). Personal epistemology and mathematics: A critical review and synthesis of research. Review of Educational Research, 74, 317–77.

Murphy, P. K., Edwards, M. N., Buehl, M. M., and Zeruth, J. A. (2007). An examination of the psychometric properties of the domain­specific beliefs questionnaire (DSBQ) with urban adolescents. Journal of Experimental Education, 76, 3–25.

Pajares, F. (1992). Teachers’ beliefs and educational research: Cleaning up a messy construct. Review of Educational Research, 62, 307–32.

Patrick, H. and Pintrich, P. R. (2001). Conceptual change in teachers’ intuitive conceptions of learning, motivation, and instructions: The role of motiv­ational and epistemological beliefs. In B. Torf and R. Sternberg (Eds.), Understanding and teaching the intuitive mind (117–43). Mahwah, NJ: Law­rence Erlbaum Associates.

Paul, R. W., Binker, A. J., Jensen, K., and Kreklau, H. (1990). Critical think-ing handbook 4th–6th grades. Dillon Beach, CA: Foundation for Critical Thinking.

What does research and theory tell us?586

Perry, W. G. Jr. (1970). Forms of intellectual and ethical development in the college years: A scheme. New York: Holt, Rinehart and Winston.

(1978). Sharing in the costs of growth. In C. A. Parker (Ed.), Encouraging development in college students (267–73). Minneapolis, MN: University of Minnesota Press.

Resnick, L. B. and Klopfer, L. W. (Eds.) (1989). Toward the thinking curric-ulum: Current cognitive research. Alexandria, VA: Association for Supervi­sion and Curriculum Development.

Shulman, L. S. (1987). Knowledge and teaching: Foundations of the new reform. Harvard Educational Review, 57, 1–22.

Stevens, R., Wineburg, S., Herrenkohl, L. R., and Bell, P. (2005). Compara­tive understanding of school subjects: Past, present, and future. Review of Educational Research, 75, 125–57.

Stiggins, R. J. (2004). Student-involved assessment FOR learning (4th edn.). Columbus, OH: Merrill Prentice Hall.

Wong, B., Khine, M. S., and Sing, C. C. (2008). Challenges and future direc­tions for personal epistemology research in diverse cultures. In M. S. Khine (Ed.), Knowing, knowledge and beliefs: Epistemological studies across diverse cultures (445–56). New York: Springer.

Winsor, D. L. (2008). Exploring preschooler’s personal epistemology using focus groups. Unpublished dissertation, University of Nevada, Las Vegas.

Zemp, L. (2009). Secondary students’ personal epistemology within the context of a poetry unit. Unpublished thesis, University of Nevada, Las Vegas.

587

ability beliefs of teachers 490–4a calling or gift 491innate 490innate and learned 491learned 491requires polish 491implications 492–4, 503–4

ability of students: teachers’ beliefs 493, 494

absolutism 7, 222, 249absolutists 66, 127, 263academic performance

and epistemic beliefs 332influence of goal orientations 334–5see also high­poverty, high­minority

studyAchievement Goals Questionnaire 447affect 101, 117, 283, 499, 508, 557Alexander, P. A. et al. 293, 395, 397, 479,

495, 505Anderson, O. R. 130–4Anderson, R. C. et al. 103Armitage, D. 537assessment of students 421–5, 457, 458,

572, 575associations 277autism 244, 252awareness 114, 412, 569

Bakken, L. et al. 37Bandura, A. 107Baron, R. M. 449Baxter Magolda, M. B. 371Beatty, P. C. 380Belenky, M. F. et al. 34, 49, 306Bempechant, J. 32Bendixen, L. D. 58, 61, 63, 64, 78,

79, 140, 165, 171, 259, 285, 416, 444

Benson, G. D. 76, 77, 78

Bentler, P. M. 354, 449Berry, J. S. 297Berthelsen, D. 473Biggs, J. B. 440Boaler, J. 299, 304, 305–7Boscolo, P. 73, 80, 184Boyes, M. 210Bråten, I. 353Brewer, W. F. 297Bronfenbrenner, U. 34, 41, 43, 56Brownlee, J. et al. 169, 473, 474, 522Buehl, M. M. et al. 293, 338, 373, 441–2,

443, 444, 446–7Burr, J. E. 69, 226, 232, 439Burrus, B. M. 335, 353Butler, R. 493

CAMCC (Cognitive­Affective Model of Conceptual Change) 261–2

Carey, S. et al. 266–7Carpendale, J. I. 229Carpenter, T. P. et al. 299certainty of knowledge 4, 265Chandler, M. J. et al. 36, 140, 210,

216, 226, 228, 229, 242, 372, 375

Chi, M. 522, 538Chinn, C. A. 267, 269citizenship 583classroom community 118classroom culture 22, 564–5, 576

ascertaining studies 305–10group work 308and mathematics­related beliefs 302,

313–15, 319psychological perspective 303social perspective 303, 304socio­constructivist perspective 302–17

classroom epistemology see epistemic climate

Index

The index was partially supported through the Publication Subvention Program of The University Of Toledo.

Index588

classroom management and organization 479

Clinchy, B. 70, 223, 224, 226, 229, 230, 232

Cobb, P. et al. 303–4, 310–12, 318–19, 426

Cognitive­Affective Model of Conceptual Change (CAMCC) 261–2

cognitive development 164, 222–3, 243–4

cognitive interviewing 369, 377–80, 389–90, 398

Cognitive Reconstruction of Knowledge Model (CRKM) 260–1

commitment in relativism 6compassion 117conceptual change 11, 258, 260–2

4th­graders 2695th­graders 269, 272–3, 277–806th­graders 2698th­graders 268–9, 270–210th­graders 277anomalous data 267–9Cognitive­Affective Model of

Conceptual Change (CAMCC) 261–2

Cognitive Reconstruction of Knowledge Model (CRKM) 260–1

educational implications 280–4epistemological resources 416influence of epistemic beliefs 267–75,

276intentionality 276research implications 285strategies 277teachers 494–9, 507

conceptual issues 556–62exosystem 555, 558–62individual system 555, 556–7macrosystem 556, 562microsystem 555, 557

Conley, A. M. et al. 267, 332connected knowing 35, 306content knowledge 482

see also domain knowledge of teachers; subject­specificity

context 517epistemic climates 84epistemological framing 414, 416,

421–5, 426epistemological resources 410, 429–31personal epistemology 517

Corno, L. 99critical reading 183critical thinking 8, 183, 282, 368

CRKM (Cognitive Reconstruction of Knowledge Model) 260–1

Cruz, J. 374culture 35, 148, 578curricular epistemology

as knowledge structures 74–5as worldviews 75–6

Daniels, D. H. 437Danovitch, J. H. 174DAP (Developmentally Appropriate

Practice) 221Depaepe, F. et al. 307, 320, 322developmental frameworks 5–9, 370

epistemological thinking 7, 222–3Intellectual and Ethical Development

6Reflective Judg ment Model 6–7,

128, 135, 168see also epistemic development

developmental psychology 164, 263–4Developmentally Appropriate Practice

(DAP) 221Dewey, J. 111Dickens, W. T. 108DiPietro, K. 48discourse patterms 68distal knowledge 187distributed knowledge 164–7, 172, 207division of cognitive labor 163–88

assessment of knowledge claims 165–7, 169, 183

distributed knowledge 164–7, 172, 207educational implications 182–6elementary students 174first­hand evaluation 169, 170justification of knowledge 168–9research implications 186–8research on epistemological beliefs

167–9second­hand evaluation 170–2source judgements 168, 176–82understanding the division 172–6

children’s understanding 173–5school support 175–6

Dobson, J. E. 75Dobson, R. L. 75documents representation 188dogmatism 375Dole, J. A. 260–1, 262domain dependency 228–43

3–5­year­olds 233definitions 228, 246domain specificity 239–41previous research 229–33relativism 233, 236, 237–8, 241

Index 589

tolerance 233, 236, 241domain knowledge of teachers

479–85classroom management and

organization 479content knowledge 482knowledge of children 482knowledge of self 483pedagogical knowledge 482implications 483, 503

domain qualities and skills of teachers 499–501

adaptability and ingenuity 499communication 499enthusiasm 500integrity and commitment 500nurturance and care 499implications 500–1

Domain­Specific Belief Questionnaire (DSBQ) 446–7

domain­specific knowledge content knowledge 482subject­specificity 84, 175–6, 178,

183–6domain specificity 22, 293–4, 397, 399,

473early childhood 239–41and epistemic climate 83–6mathematics 297well­/ill­structured domains 375, 381,

388–9, 390–2DSBQ (Domain­Specific Belief

Questionnaire) 446–7dualism 6, 222Dweck, C. S. et al. 32, 333, 335, 353,

442, 472

early childhood 220–52, 559–60, 574cognitive development 222–3,

243–4cognitive/developmental delays 244,

252Developmentally Appropriate Practice

(DAP) 221distributed knowledge 164, 173domain dependency 228–43educational context 220–2educational implications 248–52, 574egocentric subjectivity 226–8epistemic beliefs 263kindergarten 221–2, 248–51, 574pretense and belief 240realism 225–6representational diversity 211research implications 243–7

future directions 246–7

methodological considerations 243–6

social development 251–2theory of mind 129, 175, 198, 211,

223–5, 226, 238verbal ability 244–6see also epistemic development

Easter, M. 35EBI (Epistemic Belief Inventory) 382ecological system model 34, 56educational agenda 575–8

exosystem 555, 576–7individual system 555, 575–6macrosystem 556microsystem 555, 576

educational implications 23, 571conceptual change 280–4development of epistemological

thinking 215–16division of cognitive labor 182–6early childhood 248–52, 574educational model of personal

epistemology (EMPE) 82–3elementary classrooms 81epistemic beliefs 280–4epistemic development 215–16epistemological belief system 46–9epistemological framing 421, 425exosystem 555, 573–4high­poverty, high­minority study

359–61individual system 555, 571integrative model of personal

epistemology development (IM) 95–119

macrosystem 556, 574–5mathematics 320–2, 400, 401, 455–6,

458microsystem 555, 572–3scientific reasoning 149–53, 281–2students’ epistemic and ontological

cognition 398–401teacher education 457teachers’ beliefs 455–6, 457–8teaching knowledge beliefs 501–9testing 457, 458

educational model of personal epistemology (EMPE) 58–65, 61, 78, 230

application to elementary classrooms 82–3

comparison of models 63–4epistemic climate 56–8, 65–81epistemic instruction 71, 79–80epistemic knowledge representations

74–7, 80–1

Index590

Epistemological theories influencing classroom learning 61, 62–3, 64

integrative model of personal epistemology development 61, 63, 64, 78, 94–119

learners’ personal epistemology 69–71, 79

Model of Educational Reconstruction 61, 62, 63

Model of Life­Problem­Centered Pedagogy 60–2, 63

reciprocal relations 59synthesis 64–5teachers’ personal epistemology 66–9,

78–9educational psychology 264–6egocentric subjectivity 226–8Elbaz, F. 474Elby, A. 12, 57, 208, 209, 214, 429Elder, A. D. 71, 266elementary classrooms 55–89

educational implications 81educational model of personal

epistemology (EMPE) 58–65, 82–3

epistemic climate 56–8, 65–81epistemic instruction 71, 79–80epistemic knowledge representations

74–7, 80–1learners’ personal epistemology 69–71,

79reciprocal relations 77–81research implications 81teachers’ personal epistemology 66–9,

78–9elementary students 560

distribution of cognitive labour 174source judgements 176–82students’ personal epistemology 69–71,

79see also conceptual change; epistemic

beliefs; mathematics­related beliefs; scientific reasoning; students’ epistemic and ontological cognition; teachers’ beliefs: effects in mathematics

Elliot, A. J. 447embedded systemic model 34–6, 440–1EMPE see educational model of personal

epistemologyenvironment 101epistemic and ontological cognition

see students’ epistemic and ontological cognition

Epistemic and Ontological Cognition Questionnaire (EOCQ) 381–4, 389, 392, 395

Epistemic Belief Inventory (EBI) 382epistemic beliefs 258, 330–2

and academic performance 332definitions 228, 259, 270, 373development over time 331developmental psychology 263–4educational implications 280–4educational psychology 264–6and ethnicity 332factor structure 330, 331and goal orientations 335–6influence on conceptual change

267–75, 276intentionality 276research implications 284–5science education 266–7and socio­economic status 332of teachers 436younger students 263see also epistemic beliefs, goal

orientations and achievement; epistemological beliefs

epistemic beliefs, goal orientations and achievement

high­poverty, high­minority study 352–7

science learning 443–4and self­efficacy 443, 453–5theoretical framework 441–2

epistemic climate 22, 56–8, 564–5concept 56–8context­ and school subject­specificity

83–6definition 57–8developmental aspects 86–8educational model of personal

epistemology (EMPE) 82–3elementary classrooms 55–89epistemic instruction 71, 79–80epistemic knowledge representations

74–7, 80–1learners’ personal epistemology 69–71,

79literature review 65–81reciprocal relations 59, 77–81research implications 81teachers’ personal epistemology 66–9,

78–9epistemic cognition 6epistemic development 22, 129, 197–216,

559–62aesthetic judgment and personal taste

204–6

educational model (cont.)

Index 591

educational implications 215–16inter­subjective agreement 199, 215objectivity 200, 206–9, 215, 224preschoolers 164, 173, 198, 205,

211“propagated” knowledge 207social facts and shared values 201,

203–4, 208, 209–13, 215–16see also conceptual change;

developmental frameworks; early childhood; stability of knowledge; stability of teachers’ knowledge; students’ personal epistemology

epistemic dimensions see certainty of knowledge; justification of knowledge; nature of knowledge; process of knowing; source of knowledge; stability of knowledge; structure of knowledge

epistemic doubt 99, 259, 283, 557epistemic instruction

educational model of personal epistemology (EMPE) 82–3

elementary classrooms 71, 79–80empirical research 73–4impression model 72insight model 72reciprocal relations 77–81rule model 73theoretical assumptions 72–3

epistemic knowledge representations curricular epistemology as knowledge

structures 74–5curricular epistemology as worldviews

75–6educational model of personal

epistemology (EMPE) 82–3elementary classrooms 74–7, 80–1empirical research 76–7reciprocal relations 77–81

epistemic multiplier effect 100epistemic volition 99–100, 102–7

conscious or automatic? 104–5and intentionality 105–6and Schema Theory 103–4

epistemological activities 12epistemological belief system 9–10, 31,

259, 370belief dimensions 9, 32educational implications 46–9knowledge beliefs 33learning beliefs 33recursion 36–46research implications 49

epistemological beliefs definitions 259, 292–4, 519, 520

see also epistemic beliefs; epistemological framing; students’ epistemic and ontological cognition; teachers’ epistemological and ontological worldviews

epistemological dimensions see certainty of knowledge; justification of knowledge; nature of knowledge; process of knowing; source of knowledge; stability of knowledge; structure of knowledge

epistemological forms 12epistemological framing 409, 412–16

context 414, 416, 421–5, 426deliberate stability 415, 426educational implications 421, 425fostering epistemological change 417,

425–7local coherence 413methodological implications 415–16research implications 425stability 414structural stability 415, 426student assessment 421–5teacher recognition of student

approaches 416, 417–21expectation of epistemological

variability 419grain size of epistemological “unit”

420–1resources and views about learning

421Epistemological Questionnaire (EQ)

264, 372, 382epistemological resources 12, 409–12

awareness 412conceptual change 416context sensitivity 410learning 412local coherence 413nature of knowledge 411–12research implications 428–31

manipulations of context 429–31naturalistic case studies 428–9

see also epistemological framingepistemological stances 12, 520epistemological theories 10–11Epistemological theories influencing

classroom learning 61, 62–3, 64epistemological thinking 7, 215–16,

222–3epistemological worldview 519–20

see also personal epistemology; teachers’ epistemological and ontological worldviews

Index592

epistemology: definitions 518EQ see Epistemological Questionnaireequilibration 101, 110–13

Perry’s scheme 111–13Piaget’s equilibration process 110–13

Ericsson, K. A. 378ethnicity 332evaluativism 7, 67, 127, 171–2, 222, 263,

578exosystem 555, 581–3individual system 555, 579–80macrosystem 556, 583microsystem 555, 580–1

Evans, G. W. 41, 43exosystem 555

conceptual issues 558–62educational agenda 576–7educational implications 573–4evaluativism 581–3macrosystem 577–8methodological issues 565–6teacher’s role 569–70

false­belief tasks 223–4, 244, 247family systems 35Feldman, K. A. 36, 335Festinger, L. A. 111Flavell, J. H. et al. 206Flynn Effect 109Flynn, J. R. 108four­quadrant scale 523, 524–5, 527,

540–1frameworks of personal epistemology 21

developmental frameworks 5–9, 370educational model of personal

epistemology (EMPE) 82–3epistemological framing 409, 412–16scientific reasoning 145–7teaching knowledge belief framework

478Fromm, E. 212

Gadamer, H. G. 204, 210, 517Galotti, K. M. et al. 35Gava, M. 270–2Gehlbach, H. 328genetic epistemology 110Gill, M. G. et al. 67, 473goal direction 116goal orientations 328–30, 333–5, 575–6

and epistemic beliefs 335–6influence on academic performance

334–5learning goals 333performance goals 333–4

performance­approach goals 333

performance­avoidance goals 333work­avoidant goals 334see also high­poverty, high­minority

studyGoicoechea, J. 523, 526Golin, G. 74Gopnik, A. 206Graham, S. 329Greene, J. A. et al. 373, 380Greeno, J. 305–7, 306Gregoire Gill, M. 262Gregoire, M. 261–2Griffith, B. E. 76grounded theory 477Guba, E. 517Gunzenhauser, M. G. 458Guttman, L. M. 335

Haerle, F. C. 58, 61, 70, 165, 171Haes, J. 76, 78Hallett, D. et al. 382Hammer, D. et al. 12, 57, 208, 209, 214,

411, 413Harris, J. R. 108Harris, P. L. 164Hashweh, M. Z. 436Hetherington, E. M. 41high­poverty, high­minority study

336–628th­graders 344–8, 353, 354–5,

3569th­graders 348–51, 353, 355–6, 357achievement goal orientations 339educational implications 359–61epistemic beliefs 338–9epistemic beliefs, goal orientations and

achievement 352–7participants 336–8procedures 342–3research implications 358–9, 360,

361–2results 344–52study limitations 357–8

high­school students 560, 574see also conceptual change; high­

poverty, high­minority study; mathematics­related beliefs; scientific reasoning; students’ epistemic and ontological cognition

Hill, B. V. 75Hofer, B. K. 4, 10–11, 57, 60, 61, 62–3,

64, 69, 72, 79, 100, 101, 127, 132, 168, 183, 226, 232, 292, 297, 336, 373, 396, 400, 428, 438, 439, 442, 453, 518, 519, 578

Index 593

Hogan, K. 187Holt, K. 334Holt­Reynolds, D. 485Hu, L. 354

identity 304, 306–7IM see integrative model of personal

epistemology developmentindividual, education and society 23individual system 555

conceptual issues 556–7educational agenda 575–6educational implications 571evaluativism 579–80methodological issues 563–4teacher’s role 566–7

inferences 277information technology 165, 400Inhelder, B. 37integrative model of personal

epistemology development (IM) 61, 63, 64, 78, 94–119

affect 101dimensions of beliefs 100educational goals 118–19educational implications 114

students 114–16teacher support 116–18

environment 101equilibration 101, 110–13implications 102keystones 102–14mechanism of change 99–100

epistemic doubt 99epistemic volition 99–100, 102–7resolution strategies 100

meta­cognition 101Mr. Dyson’s classroom 97–8personal epistemology development

95–7personal epistemology multiplier

109reciprocal causation 100, 107–10

IQ and social factors 108–9Matthew effect 107–8Social Cognitive Theory 107social multiplier 108triadic reciprocity 107

Intellectual and Ethical Development 6intentionality 105–6, 115, 276IQ and reciprocal causation 108–9

Johnston, P. et al. 68, 73, 78Jordon, A. et al. 494justification of knowledge 4, 168–9, 265,

374, 383–4, 392, 397, 398, 489

see also epistemic dimensions; epistemological dimensions; process of knowing

Kahneman, D. 104Kalish, C. W. 177Kang, N. W. 536Karabenick, S. A. et al. 378, 385Kardash, C. M. 473Kattman, U. et al. 61, 62, 63Keil, F. C. et al. 173, 175, 177Kember, D. 538Kenny, D. A. 449Kienhues, D. et al. 185Kilbourn, B. 76Kincheloe, J. 523kindergarten 221–2, 248–51, 574King, P. 6–7, 8, 31, 97, 128, 225, 372Kitchener, K. S. 6–7, 8, 32, 97, 128, 225,

372Kitchener, R. F. 259, 374Kloosterman, P. 298, 447knowledge

beliefs 33certainty 4, 265connected knowing 35, 306as constructed 411–12, 414content knowledge 482curricular epistemology as knowledge

structures 74–5curricular epistemology as worldviews

75–6distal knowledge 187distributed knowledge 164–7, 172, 207domain knowledge of teachers

479–85inter­subjective agreement 199, 215justification 4, 168–9, 265, 374, 383–4,

392, 397, 398, 489nature of knowledge 4, 12, 140, 265,

266–7, 411–12objectivity 200, 206–9, 215, 224pedagogical knowledge 482, 486process of knowing 4, 141–2, 265as propagated stuff 207, 411–12, 413received knowing 306separate knowing 34simplicity 4, 265social/institutional facts 201source 4, 9, 12, 32, 176–82, 265,

485–9stability 9, 32, 494–9structure 9, 32, 172, 175subject­specificity 84, 175–6, 178,

183–6knowledge of children 482

Index594

knowledge of self 483knowledge revision see conceptual changeKöller, O. 301Kuhn, D. et al. 7, 8, 70, 88, 96, 124, 125,

126, 127, 128, 129, 143, 167, 171, 200, 209, 214, 215, 222, 225, 226, 229, 230, 263, 264, 332, 370, 372, 375, 397, 578, 579, 580, 583

Kwan, K. P. 538

Lam, B. H. 538Lampert, M. 296Lau, M. et al. 429learning 412

ability to learn 9, 32as accumulation 413speed of learning 9, 32

learning beliefs 33Learning Environment Preference (LEP)

133learning goals 333learning materials and procedures 118,

572see also curricular epistemology;

epistemic instruction; epistemic knowledge representations

Lee, K. 493Leggett, E. L. 333, 335, 353, 442Lent, R. W. et al. 359LEP (Learning Environment Preference)

133Lidar, M. et al. 536Liebert, R. M. 108Lincoln, Y. 517Linn, M. C. 267Lising, L. 429Louca, L. et al. 73, 80Lutz, D. J. 173

McClain, K. 310–12McGregor, H. 447macrosystem 556

conceptual issues 562educational agenda 556educational implications 574–5evaluativism 583exosystem 577–8methodological issues 566teacher’s role 570see also educational model of personal

epistemology (EMPE)Malhotra, B. A. 269Mansfield, A. 70, 223, 224, 226, 229,

230, 232Marcia, J. E. 210, 216Marra, R. 522

Mason, L. et al. 73, 80, 106, 139, 148, 184, 267–9, 270–2, 276, 277–80, 315–17

mathematical disposition 294–6educational implications 320–2heuristics methods 295knowledge base 295meta­knowledge 295positive mathematics­related beliefs

295self­regulatory skills 295

mathematics­related beliefs 292, 307ascertaining studies 305–10categories of beliefs 316and classroom culture 302, 313–15,

319classroom norms 310–12, 321–2educational implications 320–2, 400,

401epistemological beliefs 292–4identity 304, 306–7intervention studies 310–17, 320learning and performance 301–2, 438mathematical difference norm 311–12mathematical disposition 294–6mathematics learning 298maths as a domain 297past research 317–20problem­solving 298–301, 307,

313–15psychological perspective 303social perspective 303, 304, 318, 319socio­constructivist perspective 302–17students’ beliefs 296–302, 305–10,

437suspension of sense­making 299teachers’ beliefs 307see also teachers’ beliefs: effects in

mathematicsMatthew effect 107–8Measure of Intellectual Development

(MID) 133measurement of personal epistemology

23context 517four­quadrant scale 523, 524–5, 527,

540–1strategies 521, 523–7, 540students 381–4, 387–90, 396–8

Meece, J. L. 334mesosystem see educational model of

personal epistemology (EMPE)meta­cognition 101, 295, 427, 557methodological issues 562–6

exosystem 555, 565–6individual system 555, 563–4

Index 595

macrosystem 556, 566microsystem 555, 564–5

microsystem 555conceptual issues 557educational agenda 555, 576educational implications 572–3evaluativism 580–1methodological issues 564–5teacher’s role 567–9see also educational model of personal

epistemology (EMPE)MID (Measure of Intellectual

Development) 133middle school students 560, 574

epistemological framing 417–21see also conceptual change; epistemic

beliefs; high­poverty, high­minority study; mathematics­related beliefs; scientific reasoning

Middleton, M. J. 333, 334Midgley, C. et al. 333, 334, 339Milza, P. 212minority groups see high­poverty,

high­minority studyModel of Educational Reconstruction 61,

62, 63Model of Life­Problem­Centered

Pedagogy 60–2, 63modeling 117Moore, W. S. 497motivational factors 261Mottier Lopez, L. 318Muis, K. R. et al. 148, 294, 302, 307–10,

437, 439, 444, 452, 453, 455, 505multiplicity 6multiplism 7, 66, 127, 222, 263Murphy, P. K. et al. 338, 374

National Association for the Education of Young Children (NAEYC) 221

National Council for the Accreditation of Teacher Education (NCATE) 499, 500

National Council of Teachers of Mathematics (NCTM) 295, 455, 458

National Research Council 295National Science Council (NSC)

458nature of knowledge 4, 12, 140, 265,

266–7, 411–12see also epistemic dimensions;

epistemological dimensions; justification of knowledge; source of knowledge

Neber, H. 443–4

No Child Left Behind Act (2001) 221, 250, 457

Nunes, T. et al. 298

objectivity 200, 206–9, 215, 224Olafson, L. 78, 457, 474, 504ontological beliefs 246, 521, 522

see also students’ epistemic and ontological cognition

ontological training 522ontological worldview 521

see also teachers’ epistemological and ontological worldviews

ontology 520, 522–3, 526Ormrod, J. E. 43

Packer, M. 523, 526Pajares, F. et al. 334, 435paraphrasing 277Patrick, H. 473, 484, 507Pauli, C. et al. 320Paulsen, M. B. 36, 335pedagogical knowledge 482, 486performance goals 333–4

performance­approach goals 333performance­avoidance goals 333

Perkins, D. N. 125Perner, J. 211Perry, W. G. 6, 7, 8, 23, 31, 57, 111–13,

117, 127, 133, 140, 168, 225, 371, 583

personal epistemology as cognitive developmental process

222–3construct 4, 21, 520as developmental trajectory 5–9dimensions 4educational model of personal

epistemology (EMPE) 82–3as epistemological belief system 9–10as epistemological resources 12as epistemological theories 10–11integrative model of personal

epistemology development (IM) 61, 63, 64, 78, 94–119

models 58–65, 370–1, 396see also frameworks of personal

epistemology; measurement of personal epistemology; students’ epistemic and ontological cognition; teachers’ epistemological and ontological worldviews

personal epistemology multiplier 109Peterman, F. 435Phelan, J. 417

Index596

philosophical epistemology 374Piaget, J. 37, 96, 101, 110–11, 164, 211,

224Picker, S. H. 297Pines, A. L. 74Pintrich, P. R. et al. 4, 10–11, 57, 60, 100,

106, 132, 258, 276, 292, 293, 297, 439, 442, 453, 473, 484, 507

Pollock, J. L. 374Porsch, T. et al. 177poverty see high­poverty, high­minority

studypractical epistemology 187pre­reflective thinking 6preschoolers see early childhoodpretense and belief 240problem­solving 116

mathematics 298–301, 307, 313–15process of knowing 4, 141–2, 265

see also epistemic dimensions; epistemological dimensions; justification of knowledge; source of knowledge

Prosser, M. 538proximal beliefs 187Putney, L. 47

Qian, G. 335, 353quasi­reflective thinking 6

rationalism 375Ravindran, B. et al. 473realism, 127, 224, 263, 375, 526

early childhood 225–6received knowing 306reciprocal causation 100,

107–10educational model of personal

epistemology (EMPE) 82–3integrative model of personal

epistemology development (IM) 61, 63, 64, 78, 94–119

IQ and social factors 108–9Matthew effect 107–8Social Cognitive Theory 107social multiplier 108triadic reciprocity 107see also reciprocal relations

reciprocal relations 59, 77–81see also reciprocal causation

recursion 36–46, 242, 561formal operations 44–5home environments 39initial cognitive ability 39–41onset of formal operations 41–4

Redish, E. F. et al. 413, 416, 429

Reflective Judgment Model 6–7, 128, 135, 168

reflective thinking 6relativism 6, 233, 236, 237–8, 241, 523,

526Repacholi, B. M. 206resolution strategies 100Richardson, V. 472, 486Richardson, W. K. 398Rosenberg, S. et al. 418, 522Rule, D. C. 58, 61, 63, 64, 78, 79, 100,

140, 285, 416, 444Ryan, M. P. 32

Sandoval, W. A. 182, 187Scheffler, I. 72, 79Schema Theory 103–4Schoenfeld, A. H. 32, 294, 298, 299, 301,

438, 447Schommer­Aikins, M. et al. 9–10, 34,

35, 36, 37, 38, 41, 45, 49, 57, 100, 265, 331, 359, 372, 440–1, 443–4, 490

Schommer, M. 32–3, 264, 293, 330, 332, 442, 472, 519

Schraw, G. 78, 285, 453, 457, 474, 504scientific reasoning as school subject

124–535th­graders 71, 266, 272–36th­graders 134–7, 140, 141–2, 1507th­graders 266–78th­graders 137–8, 140, 141–2, 148,

152, 267, 268–9, 270–210th­graders 132–4, 140, 142–4, 150–212th­graders 130–4beliefs about nature of knowledge 140,

266–7beliefs about processes of knowing

141–2conceptual change 262, 268–9, 270–2coordination of theory and evidence

142–4educational implications 149–53,

281–2epistemic beliefs, goal orientations and

achievement 443–4epistemic framework 145–7informal contexts 125–6, 129–39modes of reasoning 130–4, 159, 160and personal epistemology 126–9,

132–4, 144, 158research implications 147–9use of theory and evidence 131–2, 157see also content knowledge; subject­

specificityScirica, F. 139, 148

Index 597

SCOPE project 282Scrivani, L. 315–17Searle, J. R. 201Seigel, H. 124self­efficacy 444, 453–5Shadish, W. R. et al. 523, 524, 539Shulman, L. S. 475, 486Shulman, V. 537Shumow, L. 437Simon, H. A. 378simplicity of knowledge 4, 265Sinatra, G. M. et al. 106, 260–1, 262,

276, 473skepticism 375Slotta, J. D. 522, 538Smith, C. L. et al. 73–4, 80Smyth (2004) 183Social Cognitive Theory 107socio­cultural development theory 47, 56socio­economic status 332Songer, N. B. 267source of knowledge 4, 9, 12, 32, 176–82,

265, 485–9see also epistemic dimensions;

epistemological dimensions; justification of knowledge; process of knowing

source of knowledge of teachers 485–9enactive experiences 487formal preparation 486formalized bodies of information 486interactive and collaborative

experiences 487observational and vicarious

experiencees 487self­reflection 487–8implications, 488, 503see also division of cognitive labor

Speers, N. 518stability of knowledge 9, 32, 494–9

see also certainty of knowledge; epistemic dimensions; epistemological dimensions; nature of knowledge; structure of knowledge

stability of teachers’ knowledge 494–9amount of change 495direction of change 496quality of change 496–7reason for change 497specific topics 497implications, 497–9, 503

Stanley­Hagan, M. 41Stanovich, K. E. 105Stathopoulou, C. 276, 277Steele, C. M. 431

Steinberg, L. et al. 38, 360Steinbring, H. 73, 80Stevens, R. et al. 176, 186Stodolsky, S. S. et al. 175, 294Strømsø, H. I. et al. 166, 353structure of knowledge 9, 32, 172, 175

see also certainty of knowledge; epistemic dimensions; epistemological dimensions; nature of knowledge; simplicity of knowledge

students’ epistemic and ontological cognition 368–401

cognitive interviewing 369, 377–80, 389–90, 398

conceptual model 373–7, 398–9dimensions 374, 382, 398dogmatism 375domain specificity 375, 381, 388–9,

390–2, 397, 399educational implications 398–401elementary and secondary students

371–3justification 374, 383–4, 392, 397,

398measurement 381–4, 387–90, 396–8personal epistemology models 370–1,

396positions 375rationalism 375realism 375research study 380–98

method 381–5data analysis 385–7results 387–95implications 395, 399–401

skepticism 375verbal protocols 378–9, 387, 397

students’ personal epistemology 114–16awareness 114early­onset/spiral­like development

69educational model of personal

epistemology (EMPE) 82–3elementary students 69–71, 79empirical findings 69–71goal direction 116intentionality 115late­onset/linear development 69learners’ personal epistemology 69–71,

79problem­solving 116reciprocal relations 59theoretical assumptions 69see also conceptual change; early

childhood; epistemic beliefs;

Index598

epistemic development; goal orientations; high­poverty, high­minority study; individual system; mathematical disposition; mathematics­related beliefs; reciprocal relations; scientific reasoning; students’ epistemic and ontological cognition

study strategies 35subject­specificity 84, 175–6, 178, 183–6

see also content knowledge; domain specificity

Tannen, D. 413teacher dispositions 499, 500teacher education 48, 284, 421, 569–70

influence of teacher beliefs 457teacher dispositions 500teaching knowledge beliefs 484–5,

488–9, 498, 507–9teacher knowledge 474–5

see also teaching knowledge belief framework; teaching knowledge beliefs

Teacher Sense of Efficacy Scale (TSES) 484

teacher support 116–18classroom community 118compassion 117learning materials and procedures 118modeling 117setting up incongruities 117

teachers’ beliefs 472about student ability 493, 494about teaching ability 490–4about teaching and learning 436constructivist epistemic beliefs 436educational model of personal

epistemology (EMPE) 82–3epistemic beliefs 436mathematics­related beliefs 307previous research 522subject­matter beliefs 261, 262see also teachers’ beliefs: effects

in mathematics; teachers’ epistemological and ontological worldviews; teaching knowledge beliefs

teachers’ beliefs: effects in mathematics 444–59

research questions 444achievement goals 444, 453–5educational implications 455–6, 457–8methodology 445

materials 446–8, 464, 465, 466

procedure 448research implications 459results 448–53self­efficacy 444, 453–5student achievement 444, 453–5students’ epistemic and learning beliefs

444teachers’ epistemological and ontological

worldviews 516alignment of beliefs and practices

537–8conflicts between beliefs and practices

532–4context 517, 540definitions 518–21, 526development of beliefs 534–6, 539epistemology and ontology 516impact of beliefs on instruction 536–7meaning and purpose 517previous research 516–18, 521–3and student engagement and

achievement 539–40study 549–50

instrument development 523–7methodology 527–9, 547–9analyses 529–32, 550–1

implications for future research 540see also teachers’ beliefs

teachers’ personal epistemology 435, 472–4

educational model of personal epistemology (EMPE) 82–3

effects on students’ beliefs 437, 438elementary teachers 66–9, 78–9see also epistemological framing;

epistemological resources; reciprocal relations; teachers’ beliefs; teachers’ beliefs: effects in mathematics; teachers’ epistemological and ontological worldviews; teaching knowledge beliefs

teacher’s role 566exosystem 555, 569–70individual system 555, 566–7macrosystem 556, 570microsystem 555, 567–9

teaching and learning Biggs’ 3P systems model 440teachers’ beliefs 436

teaching knowledge belief framework 478

ability beliefs 490–4domain knowledge 479–85domain qualities and skills 499–501source of knowledge 485–9

students’ personal epistemology (cont.)

Index 599

stability of knowledge 494–9teaching knowledge beliefs 470

personal epistemology 472–4study

methodology 476–8purpose 475

teacher knowledge 474–5teaching knowledge belief framework

478theoretical influences 471–5conclusions and implications 501–9

methodology 502–5teacher education 507–9theoretical perspective 506–7

terminology domain dependency 228, 246epistemic beliefs 228, 259, 270, 373epistemic climate 57–8epistemological beliefs 259, 292–4,

519, 520epistemological stances 12, 520epistemological worldview 519–20epistemology 518ontological beliefs 246, 521ontological worldview 521ontology 520, 526personal epistemology 4, 21, 520

testing see assessment of studentsTheory of Integrated Domains in

Epistemology (TIDE) 452theory of mind 129, 198, 211, 223–5

and autism 244, 252false­belief tasks 223–4, 244, 247and personal epistemology 224–5, 226,

238, 372theory of belief 225

think­aloud protocols 277–80thinking

critical thinking 8, 183, 282, 368epistemological thinking 7, 215–16,

222–3pre­reflective thinking 6

quasi­reflective thinking 6reflective thinking 6

TIDE (Theory of Integrated Domains in Epistemology) 452

Tobin, K. 435tolerance 233, 236, 241Triadafillidis, T. A. 455Trigwell, K. 538Tsai, C.­C. 438, 474Tschannen­Moran, M. 484TSES (Teacher Sense of Efficacy Scale)

484

VandeWalle, D. et al. 454verbal ability 244–6verbal protocols 277–80, 378–9, 387, 397Verschaffel, L. et al. 300, 313–15, 321Vosniadou, S. 105, 106, 276, 277Vygotsky, L. S. 47, 56

Wade, S. H. 72, 79Wainryb, C. et al. 230, 232, 241, 242,

251Wallace, C. S. 536Weinstock, M. 200, 215, 222, 226, 332Wellman, H. M. et al. 224West, R. F. 105Westphal, E. 60–2, 63White, B. C. 78, 474, 484Whitmire, E. 166Wildenger, L. 234Willis, G. B. 380Wilson, B. A. 535Wimmer, H. 211Wineburg, S. S. 183Wink, J. 47Woolfolk­Hoy, A. 484Wu, E. J. C. 449

Yackel, E. 319, 426Yang, F. Y. et al. 130–4Yang, O. S. 75