Developing Conceptual Understanding of Natural Selection: The Role of Interest, Efficacy, and Basic Prior Knowledge

This article was downloaded by: [Universidad Autonoma de Barcelona]On: 18 December 2014, At: 21:48Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

The Journal of Experimental EducationPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/vjxe20

Developing Conceptual Understanding ofNatural Selection: The Role of Interest,Efficacy, and Basic Prior KnowledgeLisa Linnenbrink-Garcia a , Kevin J. Pugh b , Kristin L. K. Koskey c &Victoria C. Stewart da Duke Universityb University of Northern Coloradoc The University of Akrond The University of ToledoPublished online: 13 Dec 2011.

To cite this article: Lisa Linnenbrink-Garcia , Kevin J. Pugh , Kristin L. K. Koskey & Victoria C.Stewart (2012) Developing Conceptual Understanding of Natural Selection: The Role of Interest,Efficacy, and Basic Prior Knowledge, The Journal of Experimental Education, 80:1, 45-68, DOI:10.1080/00220973.2011.559491

To link to this article: http://dx.doi.org/10.1080/00220973.2011.559491

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &

http://www.tandfonline.com/loi/vjxe20

http://www.tandfonline.com/action/showCitFormats?doi=10.1080/00220973.2011.559491

http://dx.doi.org/10.1080/00220973.2011.559491

Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14

http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/page/terms-and-conditions

THE JOURNAL OF EXPERIMENTAL EDUCATION, 80(1), 45–68, 2012Copyright C© Taylor & Francis Group, LLCISSN: 0022-0973 print/1940-0683 onlineDOI: 10.1080/00220973.2011.559491

MOTIVATION AND SOCIAL PROCESSES

Developing Conceptual Understanding of NaturalSelection: The Role of Interest, Efficacy,

and Basic Prior Knowledge

Lisa Linnenbrink-GarciaDuke University

Kevin J. PughUniversity of Northern Colorado

Kristin L. K. KoskeyThe University of Akron

Victoria C. StewartThe University of Toledo

Changes in high school students’ (n = 94) conceptions of natural selection were examined as afunction of motivational beliefs (individual interest, academic self-efficacy), basic prior knowledge,and gender across three assessments (pre, post, follow-up). Results from variable-centered analysessuggested that these variables had relatively little effect on enduring conceptual change; however,academic self-efficacy supported short-term conceptual change for girls. Results from person-centeredanalyses provided a different picture. Four profiles of motivational beliefs and basic prior knowledgewere created using hierarchical cluster analysis: (a) low interest/efficacy, low knowledge; (b) moderateinterest/efficacy, low knowledge; (c) moderate-low interest, moderate efficacy, high knowledge; and(d) high interest/efficacy, moderate knowledge. For girls, high interest and efficacy paired withmoderate basic prior knowledge (Cluster 4) resulted in the greatest conceptual change. For boys,either moderate interest and efficacy paired with high knowledge (Cluster 3) or high interest andefficacy paired with moderate knowledge (Cluster 4) resulted in the greatest enduring conceptualchange.

Keywords adolescence, conceptual change, individual differences, motivation, science education

Address correspondence to Lisa Linnenbrink-Garcia, Department of Psychology & Neuroscience, Duke University,417 Chapel Drive, Box 90086, Durham, NC 27708, USA. E-mail: [email protected]

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14

46 LINNENBRINK-GARCIA ET AL.

UNDERSTANDING CHANGE IN CONCEPTS represents a critical aspect of learning anddevelopment in science (e.g., Carey, 1999; Chi, 2008; diSessa, 2008; Vosniadou, Vamvakoussi,& Skopeliti, 2008). The issue of conceptual change has instigated a broad body of research indevelopmental psychology and science education, which have informed our basic understandingof how children’s conceptual understanding of science content develops over time and whatinstructional methodologies might be used to enhance its development. What is largely missingfrom these perspectives, however, is a consideration of students’ motivational beliefs, especiallyin combination with prior knowledge. In a seminal article, Pintrich, Marx, and Boyle (1993)argued that it was critical to infuse motivation into models of conceptual change and sciencelearning. However, there are few studies that have answered this call. As such, the purpose ofour research was to consider how two key types of motivational beliefs (interest and academicself-efficacy) combine with basic prior knowledge to shape adolescents’ conceptual change inunderstanding natural selection.

Theoretical Background

We take a cognitive-developmental perspective to investigating conceptual change. This viewassumes individuals have prior conceptions or theories regarding scientific phenomena, whichare organized in terms of frameworks, that may interfere with the learning of the scientificallyaccepted view (e.g., Chi, 2008; Vosniadou & Brewer, 1992; Vosniadou et al., 2008). Conceptuallearning thus requires an altering of prior conceptions for new conceptions to be learned. Thisview suggests that learning about concepts where students might have alternative conceptionsrequires considerable effort and a willingness to overcome prior conceptions. Given these highlevels of effort and cognitive engagement, it seems likely that motivational beliefs may facilitatethe conceptual change process (Dole & Sinatra, 1998; Lee & Anderson, 1993; Murphy & Mason,2006; Pintrich et al., 1993).

In support of this idea, Demastes, Good, and Peebles (1995) illustrated that conceptual changeis not always a rational process centered on dissatisfaction with prior conceptions coupled withviewing new conceptions as intelligible, plausible, and fruitful, as originally proposed by Posner,Strike, Hewson, and Gertzog (1982). Instead, students may alter their conceptions for motivationalreasons. For example, Demastes et al. (1995) described a student who came to understand thatthe earth had a very old age only because of her interest in and acceptance of the existence ofdinosaurs. For this student, “The world was not old because of an array of supporting evidence,but because it had to be old in order for dinosaurs to have existed” (Demastes et al., 1995, p. 660).

Recent theoretical conceptualizations of conceptual change have expanded existing modelsto incorporate motivation (Murphy & Mason, 2006; Sinatra & Mason, 2008). In Dole andSinatra’s (1998) Cognitive Reconstruction of Knowledge Model (CRKM), they argued that onemust consider the interplay of an individual’s existing conceptions and motivation, as well asthe instruction itself, in determining the individual’s type of engagement with the material andresulting level of conceptual change. With respect to motivation, they argued for the need toaccount for personal relevance, social context, and need for cognition. In CRKM, personalrelevance includes interest in the topic, emotional involvement, and high self-efficacy. In thepresent work, we focused on two aspects of personal relevance: self-efficacy and interest. Wealso considered how these motivational factors combined with basic prior knowledge to influence

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14

MOTIVATION AND CONCEPTUAL CHANGE 47

conceptual change and whether gender was a factor in the relation of these beliefs to conceptualchange.

Academic self-efficacy

Motivational researchers have long argued that academic self-efficacy, defined as students’beliefs in their capability to learn a given topic or successfully complete a particular task, isa strong predictor of engagement and learning in school (Bandura, 1993). For example, priorresearch suggests that students with higher confidence in their capabilities persist longer in theface of challenge, are more likely to use metacognitive and self-regulatory learning strategies,and perform better on academic tasks (e.g., Pintrich & Garcia, 1991; Schunk, 1991).

Given these links to engagement and learning, one would expect that academic self-efficacywould be especially beneficial for conceptual change. However, Pintrich (1999) noted that theremight be instances when academic self-efficacy undermines conceptual change. When efficacyis conceptualized as confidence in one’s understanding or knowledge, students who are overlyconfident may be more resistant to changing their prior conceptions. As such, the interplay ofprior knowledge (and whether it is accurate or inaccurate) coupled with confidence in one’sprior knowledge may alter one’s willingness to entertain alternative conceptions. Alternatively,when self-efficacy refers to confidence in one’s ability to learn, self-efficacy should enhanceconceptual change in that students will be more likely to persist even when a concept is confusingor goes against some of their prior beliefs. That is, students who are confident that they canlearn the material taught in science should be more likely to undergo conceptual change. In thepresent study, we focused on the latter possibility. As such, we assessed academic self-efficacyas students’ domain-level beliefs in their ability to learn biology.

There is preliminary evidence supporting both of these perspectives on self-efficacy. In acase study of one first grade student, Maria (1998) suggested that one student’s confidence inher knowledge initially limited the change in her conceptual understanding; however, over time,her confidence in being able to learn helped her develop a more sophisticated view of gravity.Nevertheless, caution should be used in generalizing the results from a single case study.

Individual interest

Interest also has the potential to alter individuals’ engagement in learning (e.g., Ainley, Hidi,& Berndoff, 2002; Mason & Boscolo, 2004; Schiefele, 2001, 2009), and thus should facilitateconceptual change (Murphy & Mason, 2006; Sinatra & Mason, 2008). In the present article, wedefine individual interest as an enduring predisposition toward enjoying and valuing a particulardomain (Schiefele, 2001, 2009). It includes two key components: feeling, which refers to theaffect associated with a particular domain such as liking and enjoyment, and value, which refersto the importance of a particular domain because it is useful, personally relevant, or central to theself (Schiefele, 2001, 2009).

Individual interest is thought to be long lasting and relatively stable across time and situation.Because we were interested in the interplay between motivational beliefs and basic prior knowl-edge, we did not include prior knowledge as a component of individual interest. In this sense,

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


we are not purporting to study what Hidi and Renninger (2006) called well-developed individualinterest. As with academic self-efficacy, we took a domain-level perspective rather than focusingon interest in a particular concept in order to be consistent in our investigation.

Preliminary research generally supports the idea that individual interest and conceptual changeare positively related. In several studies involving college students, Andre and Windschitl (2003)found that individual interest and prior knowledge in the domain were independent predictors ofchanges in students’ conceptual understanding of electrical circuits. An interview study conductedby Venville and Treagust (1998) to examine adolescents’ understanding of genetics suggests thispattern may be more complex. They found that adolescents who achieved high levels of conceptualchange also reported being interested in the subject. However, there were a number of studentswho had high interest but achieved relatively small amounts of conceptual change. Thus, interestalone may be a necessary but not sufficient condition for conceptual change to occur. Morerecently, Mason, Gava, and Boldrin (2008) found that children with high levels of topic interestexperienced greater levels of conceptual change regarding their conceptions of light. In addition,they found that interest was most beneficial for conceptual change when students had advancedepistemic beliefs about the nature of scientific knowledge and when they read refutational ratherthan traditional texts. Both studies highlight the need to consider how interest combines withother factors to support conceptual change.

Prior knowledge

The third variable we considered for predicting conceptual change is prior knowledge. Priorknowledge has been conceptualized in a variety of ways in the extant literature. Conceptualchange models typically focus on students’ initial levels of understanding with regard to thephenomena being studied. For example, Dole and Sinatra (1998) described prior knowledge interms of the learner’s existing conception of the topic, which includes the strength, or richnessof the conceptual understanding, coherence, and the individual’s commitment to the conception.Others have assessed prior knowledge in terms of a basic understanding of concepts relatedto the misconception rather than the misconception itself (see Qian & Alverman, 1995). Moregeneral models of domain learning differentiate between domain knowledge (e.g., subject matterknowledge) and knowledge about topics related to the domain (e.g., topic knowledge; Alexander,Jetton, & Julikowich, 1995). In the present study, we conceptualized prior knowledge as a basicunderstanding of the topic, including students’ understanding of basic concepts and vocabularyrelated to the topic. This focus on basic prior knowledge allows us to consider prior knowledge asdistinct from particular misconceptions so that we can examine how basic knowledge contributesto the change of a specific misconception. In this way, basic prior knowledge reflects students’basic understanding and overall familiarity with the topic.

There are also competing hypotheses regarding the relation of prior knowledge to conceptualchange. Some models have proposed that students with extensive experience and knowledgewithin a domain may have higher levels of commitment to that domain, thus making it moredifficult to undergo conceptual change (Dole & Sinatra, 1998). In contrast, more general modelsof domain learning suggest that the coupling of interest and prior knowledge facilitates learning(Alexander et al., 1995; Alexander & Murphy, 1998), including belief change from persuasive

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


texts (Murphy & Alexander, 2004). In this way, prior knowledge might facilitate rather thanhinder the conceptual change process, especially when combined with interest.

Gender

When considering students’ learning in science, it is also important to consider the possibilitythat boys and girls may have varying entering motivational beliefs and prior knowledge and thatthe relation of these motivational beliefs and prior knowledge to conceptual change may vary. Inthe past, researchers have focused on mean-level differences in motivation and achievement asa function of gender (for a review see Meece, Glienke, & Burg, 2006). This research generallysuggests that boys are more interested in science, have higher competency-related beliefs andachievement, and are more likely to choose science-related careers (e.g., Ceci & Williams,2007; Eccles, 1994; Halpern et al., 2007; Steinkamp & Maehr, 1984); however, in the lifesciences in particular, gender differences are reduced or even reversed (Burkam, Lee, & Smerdon,1997; Kahle, Parker, Rennie, & Riley, 1993; Steinkamp & Maehr, 1984). Given our focus onbiology, we did not expect mean-level gender differences in motivational beliefs and basicprior knowledge. However, we did expect that motivational beliefs might function differently insupporting conceptual change for boys and girls.

Given the continued under representation of women in science (Halpern et al., 2007), moti-vational beliefs such as academic self-efficacy may play a particularly important role for girls insupporting their learning and engagement in science, as girls may need especially high confidencein their ability to learn to overcome the negative stereotypes of women in science. Moreover,Pajares (1996a, 1996b) suggested that boys and girls might use different metrics to rate theirefficacy, resulting in underconfidence for girls and overconfidence for boys. Such a mismatchbetween ability and perceived competence suggests that self-efficacy may be particularly im-portant for supporting girls’ learning and achievement, especially if they are already expressingless confidence than is warranted by their performance. Similarly, self-efficacy may not be asstrong of a predictor for boys if it reflects overconfidence in their abilities. This contention thatself-efficacy may function differently by gender is supported by a study conducted by DeBackerand Nelson (1999) among high school biology students. They found that perceived ability was astrong predictor of effort, persistence, and achievement for girls; however, it was not significantlyrelated to these key learning-related outcomes for boys.

Less is known about potential gender differences in the relative influence of interest on learningand engagement. The study by DeBacker and Nelson (1999) suggested that the pattern of genderdifferences is more nuanced for the relation between interest and learning and engagement.They found that interest supported boys’ overall effort but was not related to their persistence(i.e., continued engagement even when challenged). For girls, the opposite was true. Becausepersistence may be particularly important as students work to overcome misconceptions, this maymean that interest better supports conceptual change suggesting that interest may be especiallyfacilitative of conceptual change for girls.

Overall, these findings regarding differential roles of motivational beliefs for learning and en-gagement on the basis of gender suggest that certain motivational beliefs, particularly competence-related beliefs, may be especially important for girls’ learning in biology. Accordingly, our anal-yses focused on whether the relation of motivational beliefs (academic self-efficacy, interest) to

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


conceptual change varied by gender. That is, we focused on interaction effects of gender ratherthan on main effects.

Patterns of motivational beliefs and prior knowledge

There is growing evidence that it is not enough to consider an individual’s motivation andprior knowledge in isolation; rather, it is the interplay of factors that is critical for understandingconceptual change (Murphy & Mason, 2006). For example, Alexander et al. (1995) found differentpatterns of text recall as a function of students’ prior knowledge and interest in the topic of a textpassage. Furthermore, interest and knowledge may not be sufficient for promoting conceptualchange; other motivational variables may play a role. For example, we do not know whetheradolescents who have high individual interest and high academic self-efficacy are better equippedfor the challenges associated with conceptual change than those who simply have high interestor high academic self-efficacy. It certainly seems possible that there exists a unique pairingof motivational variables that must come together in order to facilitate the conceptual changeprocess. Moreover, given the potential differences in the way that competence-related versusvalue-related beliefs relate to boys’ versus girls’ engagement and learning in biology (DeBacker& Nelson, 1999), the relation of this combination of variables to conceptual change may vary bygender.

To better understand how motivational beliefs and prior knowledge independently and syner-gistically combine to shape conceptual change, we employed a two-pronged analytical approachin the present study using variable- and person-centered analyses. The variable-centered approachallowed us to examine how, on average, varying levels of each predictor related to changes inconceptual understanding. This is the approach most frequently taken by researchers studyingconceptual change processes and therefore allowed us to build on this prior research. However,as we have previously proposed, it is also possible that the combination of variables is criticalfor shaping students’ conceptual understanding. In this way, it is not the absolute level of anyparticular variable that makes a difference in altering students’ engagement and subsequent con-ceptual understanding, but rather how these variables combine. As such, we also used clusteranalysis to create groups of students with varying levels of motivational beliefs and prior knowl-edge. This allowed the examination of differences in conceptual change as a function of varyingcombinations of motivation and prior knowledge.

Natural selection

Last, it is important to note that our investigation of conceptual change focused on the conceptof natural selection. Natural selection provides a useful context for studying the interplay ofmotivational beliefs, prior knowledge, and conceptual change for a few reasons. First, highschool students are likely to have learned about natural selection and the more general conceptof evolution in prior schooling experiences and to have had further exposure to these concepts inother informal contexts. Thus, they are likely to possess some degree of prior knowledge regardingthese concepts. In addition, research illustrates that such prior knowledge about evolution oftentakes the form of deep-seated misconceptions that are resistant to conceptual change (Bishop& Anderson, 1990; Brumby, 1984; Clough & Wood-Robinson, 1985; Evans, 2008; Ferrari &Chi, 1998; Kelemen & DiYanni, 2005; Sinatra, Brem, & Evans, 2008). Third, because of the

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


emotional characteristics associated with natural selection (Brem, Ranney, & Schnindel, 2003),it is likely that motivational factors have added significance in the conceptual change process.In line with this proposition, Demastes et al. (1995) found that motivational factors, includinginterest, played an important role in supporting or inhibiting conceptual change with respect toprinciples of evolution. Thus, it seems especially useful to study how cognitive (prior knowledge)and motivational (interest, academic self-efficacy) factors shape learning within this domain.

Present Study

Accordingly, the purpose of the present study was to investigate how patterns of motivationalbeliefs and prior knowledge related to immediate and enduring changes in adolescents’ under-standing of natural selection during a high school biology unit on evolution and to considerwhether the relation of these variables differed as a function of gender. Thus, we used a variable-and person-centered approach to investigate the direct relations of motivational beliefs (academicself-efficacy, interest) and basic prior knowledge to conceptual change and the ways in whichthese motivational beliefs combined with basic prior knowledge to influence conceptual change.Variations in the relation of the predictor variables to conceptual change as a function of genderwere explored in both sets of analyses. It is important to note that we conceptualize basic priorknowledge as distinct from students’ initial conceptions (or misconceptions) of natural selection.As such, basic prior knowledge serves as an indicator of students’ basic understanding of evolu-tion, while the pretest indicators of natural selection reflect initial conceptions/misconceptions.

For the variable-centered analyses, we examined the relation of interest, academic self-efficacy,and basic prior knowledge to immediate (posttest) and enduring (follow-up) conceptual changeand examined whether the relations varied by gender. On the basis of prior research suggestingthat interest and efficacy support learning and engagement (e.g., Schiefele, 2009; Schunk, 1991)and thus may help to support the type of deep engagement necessary for conceptual change(Dole & Sinatra, 1998), we hypothesized that biology interest and academic self-efficacy forlearning biology would facilitate conceptual change. With respect to basic prior knowledge, ouranalyses were somewhat exploratory. However, we predicted that basic prior knowledge aboutevolution (i.e., general topic knowledge rather than specific conceptions of natural selection)would either be unrelated or beneficial for conceptual change, but not inhibitory. This hypothesisis based on models of domain learning, suggesting that prior knowledge coupled with interestfacilitates learning and belief change (Alexander et al., 1995; Alexander & Murphy, 1998; Mur-phy & Alexander, 2004). However, given Dole and Sinatra’s (1998) contention that a greatercommitment to a domain among students with more extensive knowledge and experience mayhinder conceptual change, we remained open to the possibility that prior knowledge would notsupport conceptual change. Drawing from DeBacker and Nelson’s (1999) research suggestingthat interest and efficacy were more supportive of girls’ engagement and learning in high schoolbiology, we hypothesized that motivational self-beliefs would be more facilitative of girls’ con-ceptual change than boys. We did not expect a basic prior knowledge by gender interaction forconceptual change, but we tested it for exploratory purposes to examine the full range of possibleinteractions.

A second set of research questions guided the person-centered analyses. For these analyses,we did not have specific hypotheses about how the motivational beliefs and basic prior knowledgewould cluster together and therefore could not make predictions about the relation of the varying

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


profiles of motivational beliefs and basic prior knowledge to conceptual change. Rather, weframed these analyses around two guiding questions:

1. What are the underlying profiles of motivational beliefs and basic prior knowledge foradolescents in biology?

2. Do these profiles differentially predict immediate and enduring conceptual change anddo these relations vary as a function of gender?

METHOD

Approximately 1 week before the start of a 3-week science unit on evolution, adolescents com-pleted a self-report questionnaire assessing their interest in biology, self-efficacy about learningbiology, basic prior knowledge about evolution (10-item multiple choice indicator), and concep-tual understanding of natural selection (open-ended). Immediately (post) and 5 weeks (follow-up)after the end of the unit, students completed an open-ended exam question assessing their concep-tual understanding of natural selection. These data were collected as part of a larger project ex-amining students’ changing conceptions of natural selection (Pugh, Linnenbrink-Garcia, Koskey,Stewart, & Manzey, 2010); data reported here are on a subsample of participants.

Participants

Participants were 94 adolescents (M age = 15.20; 47 ninth grade, 47 tenth grade; 67% girls) whowere enrolled in 1 of 4 regular-level high school biology classes during the spring semester atan urban private Catholic high school. The classes contained ninth- and tenth-grade secondarystudents within the same classroom; thus, instruction was the same across grade levels. We foundno statistically significant differences on any of the outcome or predictor variables by grade level.Approximately three quarters of the sample was White (n = 69), with the remainder of the studentsreporting their ethnicity as African American (n = 11), Hispanic/Latino (n = 3), Asian (n = 1), andmixed or other (n = 7); three students did not report their ethnic background. All participants weretaught by the same science teacher, who had more than 20 years teaching experience. Instructionincluded direct instruction, student activities, and discussion. On a typical day, the science teacherpresented information, had students engage in an activity such as a lab or other hands-on activity,and then discussed the meaning of the activity. The teacher emphasized the importance of theactivities but did not have an explicit conceptual change, inquiry, or constructivist philosophy. Theevolution unit lasted approximately 3 weeks, 1 week focused specifically on natural selection.

Independent Variables

Motivational variables

Before the start of the biology unit, students reported on their motivational beliefs withrespect to learning about biology. In line with our theoretical approach, we focused on domain-level indicators (biology) of interest and self-efficacy rather than concept-specific measures(natural selection). Academic self-efficacy (α = .83), which assessed students’ beliefs abouttheir confidence in learning concepts in biology class (e.g., “Even if the work in biology class

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


is hard, I can learn it”), was measured using the 5-item self-efficacy scale, adapted to focus onbiology, from the Patterns of Adaptive Learning Survey (PALS; Midgley et al., 2000). As such,this indicator of self-efficacy focuses on students’ confidence in learning about biology ratherthan their confidence in their actual understanding of biology.

Interest in biology was assessed using an 8-item individual interest scale (Linnenbrink-Garciaet al., 2010) that tapped the feeling component (e.g., “I enjoy biology”) and a value/meaningcomponent (e.g., “It is important for me to learn about biology”) of interest (α = .90). All itemswere rated on a 5-point Likert-type scale.

Basic prior knowledge

Ten multiple-choice items assessed basic, factual knowledge of the broader topic of evolutionat the pretest, with items such as the following:

“Scarcity of resources, like food, and a growing population are most likely to result in . . . ”(a) homology(b) protective coloration(c) competition between individuals(d) convergent evolution(correct answer = c).

This measure was developed from the multiple-choice test typically used for the evolution unit;many of the items came from exams that were packaged with the textbook. The science instructorand an expert in high school science education reviewed all of the items to ensure that the basicconcepts pertaining to natural selection and adaptation were included. The sum of the correctresponses was used as an indicator of basic prior topic knowledge.

These items tap into basic terminology and concepts pertaining to the broader study of evolu-tion; they were not designed to assess misconceptions per se and were not specifically targeted tothe natural selection misconception. Given that students might be able to correctly identify termsor concepts but still hold deeper conceptual misconceptions, it is possible that students couldscore high on this assessment of basic prior knowledge but still have underlying misconceptionsregarding the mechanisms involved in the conceptual change process. Indeed, the basic priorknowledge measure and the pre-test assessment of natural selection conceptual understandingwere not significantly correlated (see Table 1). Thus, the assessment of basic prior knowledge canbe considered as an indicator of students’ basic understanding of evolution, reflecting a broaderbasic knowledge base.

Conceptual Understanding

A set of three open-response items administered at the pretest, posttest, and follow-up assessmentswere adapted from prior studies (Brumby, 1984; Clough & Wood-Robinson, 1985) to assessstudents’ changing conceptual understanding, including misconceptions, of natural selection. Byusing three similar versions of this question across the three time points, we were able to examinehow students’ conceptual understanding of natural selection varied as a function of motivationalbeliefs and basic prior knowledge about evolution before and after the biology unit on evolution.Three similar versions of the question for the pretest, posttest, and follow-up were developed

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


TABLE 1Bivariate Correlations and Descriptive Statistics

1 2 3 4 5 6 7

1. Interest1

2. Academic self-efficacy1 .60∗∗∗3. Basic knowledge2 .11 .20∗4. Gender3 −.08 −.12 .035. Pre conceptual understanding1 .06 .09 .15 −.076. Post conceptual understanding1 .05 .05 .21∗ .03 .047. Follow-up conceptual understanding1 .19 .23∗ .07 .04 .36∗∗∗ .37∗∗∗Overall

M 3.70 3.88 4.85 0.67 1.58 2.66 2.86SD 0.76 0.72 1.61 0.47 1.00 1.37 1.37

GirlsM 3.66 3.82 4.89 — 1.53 2.69 2.90SD 0.80 0.68 1.77 — 0.96 1.40 1.34

BoysM 3.78 4.01 4.77 — 1.68 2.60 2.77SD 0.67 0.78 1.23 — 1.09 1.33 1.45

1Scores range from 1 to 5.2Scores ranged from 0 to 10.3Boy = 0; girl = 1.∗∗∗p < .001. ∗p < .05.

in conjunction with the instructor and then pilot-tested on a small sample of approximatelyten individuals matching the target population. Pilot testing suggested that students’ responseswere similar across the three questions. All students received the same version of the item ateach assessment. These items targeted the common misconception that change over time is goaldirected and intentional such that individuals change out of need or in response to the environmentrather than change being the result of natural selection acting on a population comprised ofindividuals with varied traits (Bishop & Anderson, 1990; Evans, 2008). For example, at thepretest assessment, all students responded to the following question:

Some scientists found that most of the lizards that live in the desert are brown and most of the lizardsthat live in the forest are green. Can you explain this?How do you think this arrangement came about in the first place?What might happen if the forest became a desert over the years–say it was cut down and it rainedless? What would happen to the green lizards?

The teacher did not use the examples presented in the items during instruction.The two first authors, in collaboration with a veteran high school biology teacher, developed

a coding scheme by reviewing a random set of 15 of the conceptual change responses from thepre- and posttest assessments. The coding scheme was revised slightly after reading the full setof pretest responses, which was then recoded. The first two authors then coded all responses(interrater reliability was .91), resolving discrepancies through discussion.

The final coding scheme contained five levels. The first level, misconception, was used whenthe response conveyed the idea that individual organisms can change to fit their environment and

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


this change is a result of intentional efforts on the part of the organisms or some other processthat is not scientifically valid (e.g., “The sun in the desert turned the lizards’ skin brown”). Thesecond level, misconception with science language, reflected students’ conceptions that individualorganisms can change to fit the environment but included the use of some scientific language suchas “they will adapt” (e.g., “They adapted to their environment. . . . If the dark trees became lightthen the caterpillars would become light too”). The third level, hybrid conception was used whenresponses conveyed some understanding of natural selection, but also contained a competingidea that individual organisms can change to fit their environment (e.g., “The caterpillars learnedto blend in with their environment . . . The dark caterpillars would probably die off [if the darktree became light] because they wouldn’t be able to survive that environment”). The fourthlevel, incomplete conception, reflected responses that did not convey a clear misconception, butalso failed to provide enough detail about how adaptation occurs through natural selection. Theseresponses often included the appropriate scientific terms but did not describe the process in enoughdepth to be considered a correct conception (e.g., “The caterpillars adapted to the tree . . . If thecolor of the tree changed the dark caterpillars would be vulnerable to predators and die off”). Thefinal level, correct conception, was used for responses conveying the idea that adaptation is theresult of organisms with particular traits surviving/passing on genes while others die off/fail topass on genes (e.g., “The caterpillars that lived on the dark trees survived longer if they blended inwith their surroundings so they were dark like the trees . . . . The light ones didn’t survive as longbecause they didn’t blend in, so their light-colored genes weren’t passed on, but the dark-coloredgenes were”). To receive this highest code, the responses could not contain any misconceptionsrelated to natural selection.

The levels were assigned values as follows: misconception = 1, misconception with sciencelanguage = 1.5, hybrid conception = 3, incomplete conception = 4, and correct conception =5. The values assigned for each level of conceptual understanding were designed to reflect aninterval scale, such that the codes represented increased conceptual understanding. A value of1.5 was assigned for the second level because we perceived this level to represent only a minoradvancement over the first level. Students in this level did use the appropriate scientific terms;however, they still expressed a clear misconception regarding change in species. The third level,hybrid conception, was given a value of 3 because this level represents a midpoint between aclear misconception and a correct conception.

RESULTS

Preliminary and Descriptive Analyses

To provide a context for our analyses, we first describe the average levels of conceptual changethat occurred. Before the start of the unit, participants had relatively low levels of concep-tual understanding (M = 1.54, SE = 0.13); 80% of the participants held a clear miscon-ception at the time of pretest (score = 1 or 1.5). Students’ understanding of natural selec-tion increased during the unit (M = 2.73, SE = 0.16) and remained at similar levels in thefollow-up assessment (M = 2.70, SE = 0.16). By the time of posttest and follow-up, about25% developed a hybrid conception (score = 3) and about 35% showed no misconception

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


(score = 4 or 5). Although 40% of the participants still seemed to have misconceptions at the endof the unit, there was a clear shift in conceptual understanding as a function of the biology unit.

As part of our preliminary analyses, we also conducted independent t tests to examine mean-level differences in motivational beliefs (interest, academic-self-efficacy), basic prior knowledge,and natural selection (pre, post, follow-up) as a function of gender. There were no statisticallysignificant differences (see Table 1 for means and standard deviations).

Variable-Centered Analyses

In our first set of analyses, we conducted two hierarchical multiple regression analyses to examinethe relation of motivational beliefs and basic prior knowledge to immediate and enduring changesin students’ conceptual understanding of natural selection and to consider whether these relationsvaried by gender. Thus, we included the pretest assessment of natural selection as a control inthe first step so that we could examine change in conceptual understanding by predicting posttestand follow-up natural selection scores; motivational beliefs, basic prior knowledge, and genderwere entered in Step 2, with the gender interaction terms (Gender × Interest, Gender × Efficacy,and Gender × Basic Prior Knowledge) entered in Step 3; all nonsignificant interactions were cutfrom the final models.

Motivational beliefs, basic prior knowledge, and gender were not statistically significantpredictors of conceptual understanding at posttest or follow-up, controlling for conceptual under-standing at pretest. There was, however, a statistically significant Gender × Efficacy interactionat posttest (β = .32, sr2 = .10, p < .01).

As shown in Figure 1, girls with high academic self-efficacy learned more during the unitrelative to girls low in efficacy. The opposite pattern was observed for boys; those with lowefficacy actually learned more during the unit than those with high efficacy. This pattern suggeststhat academic self-efficacy facilitated conceptual change for girls during the biology unit, butundermined it for boys. However, this pattern dissipated at follow-up, and the only statisticallysignificant predictor of enduring conceptual understanding was students’ initial conceptions ofnatural selection (β = .35, sr2 = .12, p < .001).

Overall, the regression analyses do not provide strong support for the importance of interest,academic self-efficacy, and basic prior knowledge in predicting change in conceptual understand-ing of natural selection. However, they did suggest that, as hypothesized, academic self-efficacyfunctions differently for boys and girls. We now turn to the person-centered analyses to considerwhether the observed patterns vary when one considers how motivational beliefs and initial basicknowledge combine to shape conceptual change in biology.

Person-Centered Analyses

Our person-centered analyses focused on addressing two primary research questions:

1. How do motivational beliefs and basic prior knowledge cluster together for adolescentsin a high school biology course?

2. Does cluster group membership differentially predict immediate and enduring conceptualchange and does this relation vary as a function of gender?

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


FIGURE 1 Predicted values for Gender × Academic Self-Efficacy interaction for change in posttest conceptual under-standing. Self-efficacy was centered before entry in the regression equation. Predicted values for boys (−1) and girls (1)with high (1) and low (−1) academic self-efficacy were computed using unstandardized regression coefficients in theregression equation.

We began our analyses by first identifying the profiles. Although there are a variety ofapproaches to conducting person-centered analyses, we followed the guidelines proposed byBergman, Magnusson, and El-Khouri (2003) and used hierarchical cluster analysis to identifygroups of students with similar motivational characteristics and basic prior knowledge on thebasis of the three pretest indicators (interest, academic self-efficacy, basic prior knowledge). Thethree clustering variables (interest, self-efficacy, basic prior knowledge) were converted to z scoresbefore entry into the cluster analysis. In line with prior research using cluster analysis to createmotivational profiles (e.g., Alexander et al., 1995; Braten & Olaussen, 2005; Roeser, Strobel, &Quihuis, 2002), we used Ward’s method, which is an agglomerative procedure that begins witheach individual person as a separate cluster and then combines similar groups of people on thebasis of the sums of squares between the clusters across the clustering variables (Hair & Black,2000). Squared Euclidean differences were used as the distance measure to determine clusterformation, which tends to create clusters whose members have similar values on the clusteringvariables.

To reach a decision about the number of clusters to extract, we examined the fusion coefficientand the dendogram, which provides a representation of the cluster formation relative to thedistance between the clusters. Clusters with reasonable solutions were then extracted and therelative levels of the clustering variables within each solution were examined. On the basis ofthese procedures, a four-cluster solution seemed to best represent the data (see Table 2, Figure 2).

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


TABLE 2Mean Pretest Indicators for Interest, Efficacy, and Knowledge Across Clusters

Moderate interest/ Moderate-low interest, High interest/Low efficacy, low moderate efficacy, efficacy, moderate

Indicators (1) knowledge (2) high knowledge (3) knowledge (4) F(3, 90)

Total n 12 21 25 36Boys 4 5 5 16Girls1 8 16 19 20Interest −1.53d 0.14b −0.49c 0.76a 44.02∗∗∗Efficacy −1.60c −0.19b −0.30b 0.86a 55.32∗∗∗Knowledge −0.89c −0.83c 0.74a 0.27b 22.79∗∗∗

Note. All F values are statistically significant (p < .001). Noncommon superscripts across rows indicate means thatare significantly different using Fisher’s least significant difference (p < .05). All variables were standardized, so theoverall mean across groups is 0 (SD = 1).

1Across all cells, the sample was skewed toward girls (67%).∗∗∗p < .001.

To further support this decision, the classification feature of discriminant function analysiswas used to examine whether cluster membership could be adequately predicted through thelinear combination of the clustering variables. Using the clustering variables (interest, academicself-efficacy, and basic prior knowledge) as the predictors, we were able to accurately predictcluster membership for 86% of the cases. Boys and girls were approximately equally represented

FIGURE 2 Interest, efficacy, and basic prior knowledge by cluster group.

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


across the four clusters (χ2 = 3.85, p > .05). This suggests that cluster membership does not varyby gender.

Description of clusters

We labeled the four clusters as (1) low interest/efficacy, low knowledge (n = 12); (2) moderateinterest/efficacy, low knowledge (n = 21); (3) moderate-low interest, moderate efficacy, highknowledge (n = 25); and (4) high interest/efficacy, moderate knowledge (n = 36; see Figure 2).To examine the differences in the clustering variables for the four clusters that were formed,we conducted a multivariate analysis of variance where cluster group was the between-subjectsvariable and the three clustering variables (interest, academic self-efficacy, basic prior knowledge)were the dependent variables. At the multivariate level, there was the expected statisticallysignificant effect of cluster group: Wilks’s � = 0.13, F(9, 214.32) = 31.22, p < .001, η2 =.87. Follow-up univariate analyses showed that there were statistically significant differences ineach of the motivational variables as a function of the cluster group (see Table 2). Fischer’s leastsignificant difference (LSD) procedure was used to compare the mean-levels of the motivationaland basic prior knowledge variables across the four clusters, which are discussed in our descriptionof the clusters below.

Cluster 1: Low interest/efficacy, low knowledge

Students in the first cluster suffered from low motivational beliefs as well as low basic priorknowledge of concepts related to natural selection. They reported the lowest levels of interestand efficacy in relation to biology, which were both more than 1.5 standard deviations below themean. These levels of interest and efficacy were significantly lower than the three other clusters.Moreover, students in this cluster scored poorly on the basic prior knowledge test administered atthe pretest, with scores statistically significantly lower than Clusters 3 and 4, but not significantlylower than Cluster 2. This cluster was the smallest cluster (n = 12).

Cluster 2: Moderate interest/efficacy, low knowledge

As shown in Table 2, students in the moderate interest/efficacy, low knowledge cluster hadaverage levels of interest and efficacy for biology. However, their initial understanding of basicconcepts related to evolution was low. In comparison to the other clusters, adolescents in thiscluster began the unit with some of the lowest levels of basic prior knowledge, more than half astandard deviation below the mean and significantly lower than those in Clusters 3 and 4. Withrespect to motivation, their level of interest was about average; it was significantly higher than twoof the other clusters (Clusters 3 and 1) and significantly lower than those in Cluster 4. Students inthis cluster also had an average level of self-efficacy, which was significantly higher than those inthe low cluster (Cluster 1) but lower than those in the higher motivational cluster (Cluster 4). Assuch, we would consider the students in this cluster to represent those students of average levelsof interest and efficacy with little basic prior knowledge about evolution before the start of theunit.

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


Cluster 3: Moderate-low interest, moderate efficacy, high knowledge

The third cluster was similar to cluster two in terms of students’ motivation; however, studentsin this cluster had high levels of basic prior knowledge about evolution, which were significantlyhigher than those in the other three clusters. There were some smaller differences in motivation.In particular, the interest level in biology for students in this cluster was significantly lowerthan what was observed for Cluster 2, but still significantly higher than the interest reported forstudents in the low cluster. With levels just shy of a half of standard deviation below the mean,we would consider the interest level of this cluster to be moderate-low. Students in this clusterreported similar levels of self-efficacy as those in Cluster 2, which were significantly higher thanthose in the low cluster but lower than those in Cluster 4.

Cluster 4: High interest/efficacy, moderate knowledge

The pattern of clustering variables for the fourth and final cluster was an inverse to thatobserved in Cluster 3. Students in this cluster could be characterized as having high levels ofinterest and self-efficacy, which were significantly higher than all the other clusters. Their basicprior knowledge, on the other hand, was moderate and was significantly lower than the levelsobserved for Cluster 3. As such, Cluster 4 can be characterized as a group of students with agreat deal of interest in biology and confidence in their ability to learn biology, but only moderatelevels of basic prior knowledge.

Change in conceptual understanding as a function of clustermembership and gender

Our second research question examined whether cluster membership predicted changes inconceptual understanding at the post and follow-up assessments and whether this relation variedfor boys and girls. To this end, we conducted a repeated measures ANOVA, with gender andcluster group as between-subjects variables and time (pre, post, follow-up) as a within-subjectsvariable. As expected, there was a statistically significant change in students’ understanding ofnatural selection across the pre, post, and follow-up assessments: F(2, 166) = 27.36, p < .001,η2 = 0.33. As described earlier, students’ conceptual understanding increased during the unit.

Differences in conceptual understanding as a function of cluster membership and gender wereexamined. As expected, there was not a statistically significant main effect for gender, F(1, 83) =0.03, p > .05, across the pre, post, and follow-up assessments. Somewhat surprising is that we alsodid not find a statistically significant main effect of cluster membership: F(3, 83) = 1.93, p > .05.However, there was a statistically significant Time × Gender × Cluster group interaction, F(6,166) = 3.13, p < .01, η2 = .11 (see Figure 3), suggesting that change in conceptual understandingacross time varied on the basis of cluster group and gender. LSD post-hoc tests were used toexamine differences among the clusters for boys and girls at each of the three time points.

As shown in Figure 3, boys in all clusters increased in their conceptual understanding from preto post, but only boys in Clusters 3 (moderate-low interest, moderate efficacy, high knowledge)and 4 (high interest/efficacy, moderate knowledge) maintained or increased their conceptualunderstanding at the follow-up assessment. Indeed, while there were no statistically significantdifferences among clusters of boys at pre, F(3, 27) = 0.81, p > .05; or post, F(3, 27) = 1.22,

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


FIGURE 3 Change in conceptual understanding across time as a function of cluster membership and gender.

p > .05; there were significant differences at follow-up. In particular, boys in Clusters 3 (p = .042)and 4 (p = .032) had statistically significantly higher conceptual understanding than did boys inCluster 2 (moderate interest/efficacy, low knowledge). The difference of Clusters 3 (p = .10) and 4(p = .10) compared with Cluster 1 did not reach conventional standards for significance, but therewere marginal differences. It is interesting to note that boys in the two low-knowledge clusters(Clusters 1 and 2) lost most of the conceptual growth by the follow-up, whereas those with eitherhigh knowledge and moderate levels of motivation (Cluster 3) or high motivation and moderateknowledge (Cluster 4) were able to sustain or even increase their conceptual understanding 5weeks after the end of the unit.

A rather different pattern was observed for girls (see Figure 3). As with boys, girls did notsignificantly vary in their initial (pre) levels of conceptual understanding, F(3, 58) = 0.23, p >

.05. However, girls in Cluster 4 (high interest/efficacy, moderate knowledge) showed the greatestgains in conceptual understanding at the posttest, which was statistically significantly higher thanthose in students in the other clusters (p < .05). Girls in Cluster 4 maintained this same high levelof conceptual understanding at the follow-up, which was significantly higher than the conceptualunderstanding of girls in Cluster 3 (moderate-low interest, moderate efficacy, high knowledge,p = .02), but no longer statistically significantly different than the scores of girls in Cluster 1or 2. For girls, then, having high levels of interest and efficacy, with moderate levels of basicprior knowledge, was critical. Girls in the other three clusters showed some gain in conceptualunderstanding, but the gains were not nearly as large as those observed for Cluster 4. This is incontrast to boys who did not differ at the posttest and for whom it was equally adaptive to havehigh knowledge coupled with more moderate motivational beliefs or high motivational beliefscoupled with more moderate basic knowledge at the follow-up.

DISCUSSION

In sum, the results confirm prior theoretical positions suggesting that motivational factors areimportant to conceptual change (Dole & Sinatra, 1998; Murphy & Mason, 2006; Pintrich, 1999;Pintrich et al., 1993; Sinatra & Mason, 2008), particularly for girls. However, the results sug-gest that the relation between motivational factors and conceptual change is not simple andstraightforward for the concept of natural selection. Instead, the relation of motivational factors,

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


specifically interest and academic self-efficacy, to conceptual change varies depending on gender,basic prior knowledge, and whether immediate or enduring conceptual change is considered. Wesubsequently clarify these varied influences and discuss implications.

The variable-centered analyses suggest that self-efficacy enhances conceptual change for girls,but not boys, when it is assessed immediately after the end of an instructional unit. This is inline with the research by DeBacker and Nelson (1999) on students’ engagement and learningin high school biology and supports Pintrich’s (1999) hypothesis that students’ confidence intheir ability to learn in a particular domain can either support or undermine conceptual change,although Pintrich did not suggest that these effects would vary on the basis of gender. Somewhatsurprisingly and counter to our hypotheses, the variable-centered analyses also suggest thatneither interest nor basic prior knowledge was related to immediate or enduring conceptualchange. Moreover, the observed Gender × Self-Efficacy interaction was not maintained at thefollow-up. This inconsistent pattern of findings for the Gender × Self-Efficacy interaction atthe posttest and follow-up assessments suggests that this pattern may not be robust and futureresearch needs to investigate whether these results can be replicated. However, as subsequentlydiscussed in greater detail, our person-centered analyses present a somewhat different picturesuggesting that interest is related to conceptual change in science, but that the relation varies asa function of self-efficacy, basic prior knowledge, and gender. Moreover, these person-centeredanalyses further suggest that gender interactions are important at the follow-up when self-efficacy,interest, and basic prior knowledge are all taken into account.

As such, we now turn to a discussion of the more nuanced pattern of findings revealed bythe person-centered analyses. To simplify the discussion of these results, we focus on enduringconceptual change related to natural selection, as enduring conceptual understanding is theprimary concern of science educators. These results suggest that interest and self-efficacy areboth important to changes in conceptions of natural selection; however the relation also dependson gender and the level of basic prior knowledge. For boys, it appears that possessing high interestand self-efficacy is important for promoting enduring conceptual change if boys only posses amoderate level of basic prior knowledge. However, for those boys who already have high basicprior knowledge, only average or moderate levels of interest and self-efficacy are needed to reachreasonable levels of enduring conceptual understanding.

In contrast, it appears that, for girls, possessing high interest and self-efficacy is necessaryfor enduring conceptual change in biology, and that this high interest and self-efficacy mustbe paired with at least moderate basic prior knowledge. Thus, unlike the boys who could makeup for more average levels of efficacy and interest when they had high basic prior knowledge,only those girls with high levels of interest and efficacy seemed to experience and maintain thegreatest conceptual gains. This suggests that while interest or efficacy alone might be sufficientfor many types of learning, conceptual change in biology requires a deeper level of engagement,that, at least for girls, may only be realized when the perfect storm of motivational beliefscombine with at least some basic prior knowledge in the domain.

It is especially interesting that the supportive patterns of motivational beliefs and prior knowl-edge for boys and girls suggest that there may be different mechanisms for facilitating conceptualchange in science among boys and girls. One possible explanation for these different patternsis that girls may conceptualize self-efficacy in a different way. Indeed, Pajares (1996a, 1996b)pointed out that girls may be using a different metric in rating self-efficacy than boys such thatboys are overconfident in their judgments of competence (or girls are underconfident). It is also

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


possible that boys rate self-efficacy in a different way—perhaps focusing more on their confidencein their knowledge, while girls may focus more on their ability to learn. If this is the case, thefinding from the multiple regression analyses (see Figure 1) in which high self-efficacy supportedconceptual change for girls but undermined it for boys aligns well with Pintrich’s (1999) proposalthat self-efficacy could either support or undermine conceptual change.

With respect to the cluster patterns, it may be that girls need the added support of interest,self-efficacy, and prior knowledge to support persistence and eventual conceptual change inbiology. DeBacker and Nelson (1999) reported that interest predicted persistence for girls whileperceived-ability predicted achievement. Thus, to reach the high-level engagement needed forconceptual change, girls might benefit more from having interest and self-efficacy. Basic priorknowledge could also play a role here by providing girls with additional confidence in theirbasic understanding of biology so that they are willing to struggle with considering alternativeconceptions. Future research will need to further explore these differences to see if they canbe replicated in another sample and to better understand the potential reasons for these varyingpatterns by gender.

Limitations and Future Directions

The present study provides preliminary evidence that various configurations of motivationalbeliefs and basic prior knowledge relate to changes in conceptual understanding in science andthat these relations vary by gender. As cluster analysis can be sensitive to the sample on which theclusters are developed (Bergman et al., 2003; Hair & Black, 2000), future research should replicatethe person-centered analyses and examine whether similar patterns emerge for conceptual change.In addition, given the relatively small sample, the analyses comparing boys and girls should bestudied in a larger sample to see if similar patterns emerge. This is especially important, as therewere only 4–5 boys in some of the clusters in this study. Further, future research should replicatethis study in other school contexts such as public school settings.

It would also be beneficial to replicate this study in the context of students receiving instruc-tion designed to facilitate conceptual change. In line with other studies of natural selection (e.g.,Bishop & Anderson, 1990; Brumby, 1984), we observed that many individuals held onto existingmisconceptions either in part or in full. One consequence of the generally modest level of con-ceptual change was a lessening of variability between individuals on the posttest and follow-uplearning measures, thus reducing the probability of observing statistically and practically sig-nificant differences. The patterns observed in this study may be more pronounced in a sampledemonstrating greater conceptual change. Hence, we recommend that follow-up studies employ-ing similar person-centered analysis be conducted in the context of students receiving effectiveconceptual change instruction.

Last, it is important to keep in mind that learning about natural selection may present uniquechallenges (Griffith & Brem, 2004). Given the emotionally charged nature of learning aboutevolution (Brem et al., 2003), it is possible that the particular pattern of findings observed forlearning about natural selection may not replicate when other less controversial topics are studied.Therefore, future research should examine whether motivational beliefs combine in the same waywith basic prior knowledge and gender in other domains. It is important to keep in mind that thegender differences we observed occurred in relation to students’ learning in biology. We make no

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


claims that these patterns would appear in other domains, especially those outside of the naturalsciences or even in the physical sciences (see Burkam et al., 1997).

Implications

Our results suggest that person-centered analyses are important as a compliment to variable-centered analyses in studying factors related to conceptual change. Person-centered analysesallow insight into how variables combine to shape conceptual change. Such effects can be missedin variable-centered analyses.

Drawing from these person-centered analyses, our results suggest that the combination ofmotivational beliefs and basic prior knowledge is critical for supporting conceptual change inscience. Boys and girls undergo enduring conceptual change when they have high interest andacademic self-efficacy, coupled with at least moderate levels of basic prior knowledge. Thissuggests that it is not only important to facilitate motivation, as suggested by Pintrich et al.’s(1993) earlier calls for infusing motivation into cold models of conceptual change, but that onemust also consider how to foster at least moderate levels of basic prior knowledge. Moreover,high self-efficacy and interest may be particularly critical for girls in science.

The implications of these findings for practice are fairly straightforward. Teachers can likelyfoster conceptual change by developing basic prior knowledge, self-efficacy, and interest. Basicprior knowledge can be developed by designing learning progressions around core concepts. Insuch learning progressions, instruction is designed to engage students with the central ideas of adiscipline and to support students’ development of successively deeper understanding of these coreconcepts across grade levels (Duschl, Schweingruber, & Shouse, 2007). In terms of supportingacademic self-efficacy, providing feedback that emphasizes that past success is related to efforthelps to support competence-related beliefs (Schunk, 1982). In addition, teachers’ (and parents’)expectancies and beliefs about students’ abilities are also related to competence-related beliefs(Eccles, 1983). Making teachers aware of possible underlying biases regarding girls’ learningin science and helping them to develop strategies to support girls’ competence-related beliefsmay be especially important for enhancing girls’ competency beliefs (Eccles, 1994; Halpernet al., 2007; Haussler & Hoffman, 2002). Strategies for supporting interest development can bedrawn from work on situational interest (Bergin, 1999; Mitchell, 1993; Hidi & Renninger, 2006;Schiefele, 2009). Such work suggests that interest may be promoted by providing students withchoices and supporting autonomy, helping to make clear ties between the course material andeveryday life, enhancing students’ feelings of belongingness and connection with the teacher, andcreating opportunities for students to be actively involved in learning. Supporting belongingnessmay be especially important for girls, as girls often view science as a male domain (Baker &Leary, 1995) and may feel less connected to the domain.

Conclusion

Overall, the findings from the present study suggest that adolescents’ motivational beliefs andbasic prior knowledge are important predictors of their change in conceptions of natural selection.Students’ who enjoy and find value in studying biology, who are confident that they can learnthe concepts taught in biology, and who have at least moderate levels of basic knowledge aboutnatural selection are more likely to develop an advanced and enduring conceptual understanding

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


of natural selection. There do, however, appear to be gender differences in these patterns ofbeliefs, suggesting that high levels of interest and academic self-efficacy are especially importantfor supporting girls’ conceptual change in science. As such, our results suggest that educatorsshould work toward creating classroom environments that help support interest and academic self-efficacy, especially among girls. This, coupled with a sequencing of instruction so that studentshave some background knowledge in the domain, should help to facilitate conceptual change inhigh school science classrooms.

AUTHOR NOTES

This research was supported by a grant from the Jacobs Foundation to the first two authors. Thefindings and views reported in this manuscript are the authors’ and do not necessarily reflect theviews of the Jacobs Foundation. An earlier version of this article was presented at the biannualmeeting of the Society for Research on Child Development.

Lisa Linnenbrink-Garcia earned her Ph.D. in education and psychology from the Universityof Michigan, Ann Arbor, and is currently an assistant professor of psychology and education atDuke University. Dr. Linnenbrink-Garcia studies the development of achievement motivation andthe interplay among motivation and emotion in shaping students’ learning and engagement inschool settings. Kevin J. Pugh is an associate professor at the University of Northern Colorado.He earned his Ph.D. in educational psychology from Michigan State University. Dr. Pugh studiesthe intersection of motivation and transfer with a particular interest in how science learning canbe transformative. Kristin L. K. Koskey is an assistant professor of education at The Universityof Akron, where she teaches research design and assessment. She earned her Ph.D. in educationalresearch and measurement from The University of Toledo. Dr. Koskey’s research interests includepsychometrics, applications of the Rasch model to construct measures of educational constructs,and mixed methods research. Victoria C. Stewart is an assistant professor of education at TheUniversity of Toledo, where she also earned her Ph.D. in curriculum and instruction. Dr. Stewart’sresearch interests include studying individual’s experience of academic content/contexts, with anemphasis on the development of interest and the development of measures to assess educationalconstructs using the Rasch model.

REFERENCES

Ainley, M., Hidi, S., & Berndorff, D. (2002). Interest, learning and the psychological processes that mediate theirrelationship. Journal of Educational Psychology, 94, 545–561. doi:10.1037/0022-0663.94.3.545

Alexander, P. A., Jetton, T. L., & Kulikowich, J. M. (1995). Interrelationship of knowledge, interest, and recall: Assessinga model of domain learning. Journal of Educational Psychology, 87, 559–575. doi:10.1037/0022-0663.87.4.559

Alexander, P. A., & Murphy, P. K. (1998). Profiling the differences in students’ knowledge, interest, and strategicprocessing. Journal of Educational Psychology, 90, 435–447. doi:10.1037/0022-0663.90.3.435

Andre, T., & Windschitl, M. (2003). Interest, epistemological belief, and intentional conceptual change. In G. M. Sinatra& P. R. Pintrich (Eds.), Intentional conceptual change (pp. 173–197). Mahwah, NJ: Erlbaum.

Baker, D., & Leary, R. (1995). Letting girls speak out about science. Journal of Research in Science Teaching, 32, 3–27.doi:10.1002/tea.3660320104

Bandura, A. (1993). Perceived self-efficacy in cognitive development and functioning. Educational Psychologist, 28,117–148. doi:10.1207/s15326985ep2802 3

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


Bergin, D. A. (1999). Influences on classroom interest. Educational Psychologist, 34, 87–98. doi:10.1207/s15326985ep3402 2

Bergman, L. R., Magnusson, D., & El-Khouri, B. M. (2003). Studying individual development in an interindividualcontext: A person-oriented approach. Mahwah, NJ: Erlbaum.

Bishop, B. A., & Anderson, C. W. (1990). Student conceptions of natural selection and its role in evolution. Journal ofResearch in Science Teaching, 27, 415–427. doi:10.1002/tea.3660270503

Braten, I., & Olaussen, B. S. (2005). Profiling individual differences in student motivation: A longitudinalcluster-analytic study in different academic contexts. Contemporary Educational Psychology, 30, 359–396.doi:10.1016/j.cedpsych.2005.01.003

Brem, S. K., Ranney, M., & Schindel, J. (2003). Perceived consequences of evolution: College students perceive negativepersonal and social impact in evolutionary theory. Science Education, 87, 181–206. doi:10.1002/sce.10105

Brumby, M. N. (1984). Misconceptions about the concept of natural selection by medical biology students. ScienceEducation, 68, 493–503. doi:10.1002/sce.3730680412

Burkam, D. T., Lee, V. E., & Smerdon, B. A. (1997). Gender and science learning early in high school: Subject matterand laboratory experiences. American Educational Research Journal, 34, 297–331. doi:10.2307/1163360

Carey, S. (1999). Sources of conceptual change. In E. K. Scholnick, K. Nelson, S. A. Gelman, & P. H. Miller (Eds.),Conceptual development: Piaget’s Legacy (pp. 293–326). Mahwah, NJ: Erlbaum.

Ceci, S. J., & Williams, W. M. (2007). Why aren’t more women in science? Top researchers debate the evidence.Washington, DC: American Psychological Association.

Chi, M. T. H. (2008). Three types of conceptual change: Belief revision, mental model transformation, and categoricalshift. In S. Vosniadou (Ed.), International handbook of research on conceptual change (pp. 61–82). New York, NY:Routledge.

Clough, E. E., & Wood-Robinson, C. (1985). How secondary students interpret instances of biological adaptation. Journalof Biological Education, 19, 125–130.

DeBacker, T. K., & Nelson, R. M. (1999). Variations on an expectancy-value model of motivation in science. ContemporaryEducational Psychology, 24, 71–94. doi:10.1006/ceps.1998.0984

Demastes, S. S., Good, R. G., & Peebles, P. (1995). Students’ conceptual ecologies and the process of conceptual changein evolution. Science Education, 79, 637–666. doi:10.1002/sce.3730790605

diSessa, A. A. (2008). A bird’s-eye view of the “pieces” vs. “coherence” controversy (from the “pieces” side of thefence). In S. Vosniadou (Ed.), International handbook of research on conceptual change (pp. 35–60). New York, NY:Routledge.

Dole, J. A., & Sinatra, G. M. (1998). Reconceptualizing change in the cognitive construction of knowledge. EducationalPsychologist, 33, 109–128. doi:10.1207/s15326985ep3302&3 5

Duschl, R. A., Schweingruber, H. A., & Shouse, A. W. (2007). Taking science to school: Learning and teaching sciencein grades K-8. Board on Science Education, Center for Education, Division of Behavioral and Social Sciences andEducation. Washington, DC: The National Academies Press.

Eccles, J. S. (1983). Expectancies, values, and academic behaviors. In J. T. Spence (Ed.), Achievement and achievementmotives (pp. 75–146). San Francisco, CA: Freeman.

Eccles, J. S. (1994). Understanding women’s educational and occupational choices: Applying the Eccles et al. model ofachievement-related choices. Psychology of Women Quarterly, 18, 585–609. doi:10.1111/j.1471-6402.1994.tb01049.x

Evans, E. M. (2008). Conceptual change and evolutionary biology: A developmental analysis. In S. Vosniadou (Ed.),International handbook of research on conceptual change (pp. 263–294). New York, NY: Routledge.

Ferrari, M., & Chi, M. T. H. (1998). The nature of naive explanations of natural selection. International Journal of ScienceEducation, 20, 1231–1256.

Griffith, J. A., & Brem, S. K. (2004). Teaching evolutionary biology: Pressures, stress, and coping. Journal of Researchin Science Teaching, 41, 791–809. doi:10.1002/tea.20027

Hair, J. F., & Black, W. C. (2000). Cluster analysis. In L. G. Grimm & P. R. Yarnold (Eds.), Reading and understandingmore multivariate statistics (pp. 147–205). Washington, DC: American Psychological Association.

Halpern, D. F., Benbow, C. P., Geary, D. C., Gur, R. C., Hyde, J. S., & Gernsbache, M. A. (2007). The science of sexdifferences in science and mathematics. Psychological Science in the Public Interest, 8, 1–51. doi:10.1111/j.1529-1006.2007.00032.x

Haussler, P., & Hoffmann, L. (2002). An intervention study to enhance girls’ interest, self-concept, and achievement inphysics classes. Journal of Research in Science Teaching, 39, 870–888. doi:10.1002/tea.10048

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


Hidi, S., & Renninger, K. A. (2006). The four-phase model of interest development. Educational Psychologist, 41,111–127. doi:10.1207/s15326985ep4102 4

Kahle, J. B., Parker, L. H., Rennie, L. J., & Riley, D. (1993). Gender differences in science education: Building a model.Educational Psychologist, 28, 379–404. doi:10.1207/s15326985ep2804 6

Kelemen, D., & DiYanni, C. (2005). Intuitions about origins: Purpose and intelligent design in children’s reasoning aboutnature. Journal of Cognition and Development, 6, 3–31. doi:10.1207/s15327647jcd0601 2

Lee, O., & Anderson, C. W. (1993). Task engagement and conceptual change in middle school science classrooms.American Educational Research Journal, 30, 585–610. doi:10.2307/1163304

Linnenbrink-Garcia, L., Durik, A. M., Conley, A. M., Barron, K. E., Tauer, J. M., Karabenick, S. A., & Harackiewicz,J. M (2010). Measuring situational interest in academic domains. Educational and Psychological Measurement, 70,647–671. doi:10.1177/0013164409355699

Maria, K. (1998). Self-confidence and the process of conceptual change. In B. Guzzetti & C. Hynd (Eds.), Perspectiveson conceptual change (pp. 7–16). Mahwah, NJ: Erlbaum.

Mason, L., & Boscolo, P. (2004). Role of epistemological understanding and interest in interpreting acontroversy and in topic-specific belief change. Contemporary Educational Psychology, 29, 103–128.doi:10.1016/j.cedpsych.2004.01.001

Mason, L., Gava, M., & Boldrin, A. (2008). On warm conceptual change: The interplay of text, epistemological beliefs,and topic interest. Journal of Educational Psychology, 100, 291–309. doi:10.1037/0022-0663.100.2.291

Meece, J. L., Glienke, B. B., & Burg, S. (2006). Gender and motivation. Journal of School Psychology, 44, 351–373.doi:10.1016/j.jsp.2006.04.004

Midgley, C., Maehr, M. L., Hruda, L. Z., Anderman, E., Anderman, L., Freeman, K. E., . . . , Urdan, T. (2000). Manualfor the Patterns of Adaptive Learning Scales (PALS). Ann Arbor, MI: University of Michigan.

Mitchell, M. (1993). Situational interest: Its multifaceted structure in the secondary school mathematics classroom.Journal of Educational Psychology, 85, 424–436. doi:10.1037/0022-0663.85.3.424

Murphy, P. K., & Alexander, P. A. (2004). Persuasion as a dynamic, multidimensional process: An investi-gation of individual and intraindividual differences. American Educational Research Journal, 41, 337–363.doi:10.3102/00028312041002337

Murphy, P. K., & Mason, L. (2006). Changing knowledge and changing beliefs. In P. A. Alexander & P. Winne (Eds.),Handbook of educational psychology (2nd ed., pp. 305–324). New York: Erlbaum.

Pajares, F. (1996a). Self-efficacy beliefs and mathematical problem-solving of gifted students. Contemporary EducationalPsychology, 21, 325–344. doi:10.1006/ceps.1996.0025

Pajares, F. (1996b). Self-efficacy beliefs in academic settings. Review of Educational Research, 66, 543–578.doi:10.2307/1170653

Pintrich, P. R. (1999). Motivational beliefs as resources for and constraints on conceptual change. In W. Schnotz, S.Vosniadou, & M. Carretero (Eds.), New perspectives on conceptual change (pp. 33–50). Oxford, England: ElsevierScience.

Pintrich, P. R., & Garcia, T. (1991). Student goal orientation and self-regulation in the college classroom. In M. L. Maehr& P. R. Pintrich (Eds.), Advances in motivation and achievement (Vol. 7, pp. 371–402). Greenwich, CT: JAI Press.

Pintrich, P. R., Marx, R., & Boyle, R. (1993). Beyond cold conceptual change: The role of motivational beliefs andclassroom contextual factors in the process of conceptual change. Review of Educational Research, 63, 167–199.doi:10.2307/1170472

Posner, G. J., Strike, K. A., Hewson, P. W., & Gertzog, W. A. (1982). Accommodation of a scientific conception: Towarda theory of conceptual change. Science Education, 66, 211–227. doi:10.1002/sce.3730660207

Pugh, K. J., Linnenbrink-Garcia, L., Koskey, K. L. K., Stewart, V. C., & Manzey, C. (2010). Motivation, learning, and trans-formative experience: A study of deep engagement in science. Science Education, 94, 1–28. doi:10.1002/sce.20344

Qian, G., & Alverman, D. (1995). Role of epistemological beliefs and learned helplessness in secondary school stu-dents’ learning science concepts from text. Journal of Educational Psychology, 87, 282–292. doi:10.1037/0022-0663.87.2.282

Roeser, R. W., Strobel, K. R., & Quihuis, G. (2002). Studying early adolescents’ academic motivation, social-emotionalfunctioning, and engagement in learning: Variable- and person-centered approaches. Anxiety, Stress, & Coping, 15,345–368. doi:10.1080/1061580021000056519

Schiefele, U. (2001). The role of interest in motivation and learning. In J. M. Collis & S. Messick (Eds.), Intelligence andpersonality: Bridging the gap in theory and measurement (pp. 163–194). Mahwah, NJ: Erlbaum.

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14


Schiefele, U. (2009). Situational and individual interest. In K. R. Wentzel & A. Wigfield (Eds.), Handbook of motivationat school (pp. 197–222). New York, NY: Routledge.

Schunk, D. H. (1982). Effects of effort attributional feedback on children’s perceived self-efficacy and achievement.Journal of Educational Psychology, 74, 548–556. doi:10.1037/0022-0663.74.4.548

Schunk, D. H. (1991). Self-efficacy and academic motivation. Educational Psychologist, 26, 207–231.doi:10.1207/s15326985ep2603&4 2

Sinatra, G. M., Brem, S. K., & Evans, E. M. (2008). Changing minds? Implications of conceptual change for teaching andlearning about biological evolution. Evolution: Education and Outreach, 1, 189–195. doi:10.1007/s12052-008-0037-8

Sinatra, G. M., & Mason, L. (2008). Beyond knowledge: Learner characteristics influencing conceptual change. In S.Vosniadou (Ed.), International handbook of research on conceptual change (pp. 560–582). New York, NY: Routledge.

Steinkamp, M. W., & Maehr, M. L. (1984). Gender differences in motivational orientations toward achievement in schoolscience: A quantitative synthesis. American Educational Research Journal, 21, 39–59. doi:10.2307/1162353

Venville, G. J., & Treagust, D. F. (1998). Exploring conceptual change in genetics using a multidimensionalinterpretive framework. Journal of Research in Science Teaching, 35, 1031–1055. doi:10.1002/(SICI)1098-2736(199811)35:9<1031::AID-TEA5>3.0.CO;2-E

Vosniadou, S., & Brewer, W. F. (1992). Mental models of the earth: A study of conceptual change in childhood. CognitivePsychology, 24, 535–585. doi:10.1016/0010-0285(92)90018-W

Vosniadou, S., Vamvakoussi, X., & Skopeliti, I. (2008). The framework theory approach to the problem of conceptualchange. In S. Vosniadou (Ed.), International handbook of research on conceptual change (pp. 3–34). New York, NY:Routledge.

Dow

nloa

ded

by [

Uni

vers

idad

Aut

onom

a de

Bar

celo

na]

at 2

1:48

18

Dec

embe

r 20

14

Documents

Developing Conceptual Understanding of Natural Selection: The Role of Interest, Efficacy, and Basic Prior Knowledge