Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
DOCUMENT RESUME
ED 373 344 CS 214 468
AUTHOR Baer, Matthias; And OthersTITLE How Do Expert and Novice Writers Differ in Their
Knowledge of the Writing Process and Its Regulation(Metacognition) from Each Other, and What Are theDifference in Metacognitive Knowledge betweenWriters of Different Ages?
PUB DATE Apr 94CONTRACT 11-27673-89NOTE 44p.; Paper presented _ _..: Annual Meeting of the
American Educational Research Association (75th, NewOrleans, LA, April 4-8, 1994) and the EuropeanAssociation for Research on Learning and InstructionConference (5th, Aix-en-Provence, France, August31-September 5, 1993).
PUB TYPE Speeches/Confe7ence Papers (150) ReportsResearch /Technical (143)
EDRS PRICE MF01/PCO2 Plus Postage.DESCRIPTORS Adolescents; AdJlts; Comparative Analysis; Foreign
Countries; Inte....mediate Grades; *Metacognition;Secondary Education; Writing (Composition); WritingAchievement; *Writing Processes; Writing Research
IDENTIFIERS Collaborative Writing; Switzerland
ABSTRACTA study diagnosed the nature of different text
production abilities of Swiss grade 5 and grade 9 students, and ofadults, and diagnosed good and poor writers' differences within eachof these three age groups. Subjects, 36 in all, were uniformlydivided between 11- and 15-year-cdd students from the Bernesesecondary school and high school teachers. The 12 subjects of eachage group were made up of six good and six poor text writers. Eachsubject worked with an additional person as a pair. Each pair wasgiven the task of writing an article about the art and personage of awell-known Swiss iron-sculptor. The "tutor" guided the "tutee"through the task of writing the text. The tutee had to follow thetutor's advice and w., responsible for the formulation and writingdown of the surface structure sentences. Subjects were videotaped asthey wrote the text. Results indi.:ated that the most dominant factorfor the differences among subject's was the age factor. On the whole,the subjects from grade 5 and grade 9 were generally far more similarthan expected. Grade 9 subjects, who have almost finished school, didnot seem to possess a well- developed metacognitive knowledge base forwriting compared to that of the adult subjects. (Contains 17 figuresillustrating various aspects of data gathering, correlation analyses,and the model underlying the research.) (RS)
**************************** ******************************
Reproductions supplied by EDRS are the best that can be madefrom the or..ginal document.
*********************************W*************************************
r6
OHP Transparency 0: Title of Contribution with Names and InstitutionalAffiliation
How Do Expert and Novice Writers Differ in Their Knowledge of the Writing
Process and Its Regulation (IVIetacogni.00n) From Each Other, and What Are
the Differences in Metacognitive Knowledge Between Writers of Different
Ages? 1
Matthias BAER, Armin HOLLENSTEIN, Margaretha HOFSTETTER, Michael FUCHS &
Monika REBER-WYSS, Department of Pedagogical Psychology, University of Berne,
Switzerland 2
Abstract:This study is a diagnosis of the nature of different text production abilities of grade 5 and
grade 9 students, and of adults. It is further a diagnosis of good and poor writers'
differences within each of these three age groups. More specifically, this study
investigates the following: (1) How do younger and older writers differ from each other in
their metacognitive knowledge of writing? (2) What are the differences in metacognitive
knowledge between good and poor writers in each of the three Ege groups? The study
also gives (3) an idea of how the metacognitive knowledge of writing develops from
grade 5 to grade 9 and in adult writers. In addition, it provides (4) insic. into the
understanding and management of the writing processes of regular school students.
Results of this kind should be considered in intervention programs aimed at improving
students' writing abilities.U S. DEPARTMENT Of EDUCATION
Ofl,ce of Educattonal Research and Improvement
EDUCATIONAL RESOURCES INFORMATIONCENTER (ERIC)
This document has been reproduced asrecynyed Iron the person or otganizationortginafing it
Cl Minor changes bane bean made to improvereproduCbon quality
PERMISSION TO REPRODUCE THISMATERIAL HAS BEEN GRANTED BY
Points of new or opinions stated in this document do not necessarily represent official TO THE EDUCATIONAL RESOURCES
Overview: OERI Poston or POI,CY IN cOHMAT ION CENTER (ERIC)
The study is part of a three year research project which is supported by the Swiss
National Science Foundation (Grant # 11-27673.89). It is a cognitive-psychological study
and deals with metacognitive aspects of writing. The aim of the project is:
OHP Transparency 1: The Three Parts of the Project
C--/1 1 Paper presented at 5th Conference of European Association for Research on Learning and Instruction, Aix-en-Provence, France, August 31 - September 5, 1993.
2 The authors gratefully acknowledge the contribution of Erja BAER-HEIKKILA, Sam HIRSBRUNNER, MartinSTADELMANN, and Susanne RIDEGG.
2 BEST COPY AVAILABLE
1 to further develop a constructive structural model of writing (theoretical part);
2) to analyse the production process related cognitive and metacognitive abilities of
writing which school students of different ages and adults have (analytical part),
and
3) on the basis of 1) and 2) to develop a cognitive and metacognitive intervention
program to foster students' ability to write texts (intervention part).
Today, we concentrate on the analytical part of the study. In as far as it is necessary tounderstand this, we will also deal with the theoretical part. The intervention will not be
dealt with. In the analytical part of the project we investigated the following two questions:
OHP Transparency 2: The Two Questions for Analysis
1) How do younger and older writers differ in their self-regulatory knowledge of thewriting process (metacognition) and in their knowledge of the characteristics of textproducts (text organization)? This is what we call diachronical comparison.
2) How do aood and poor writers differ in their self-regulatory knowledge of the textproduction process and in their knowledge of the characteristics of text products?This is what we call synchronical comparison.
The investigation is therefore based on a 2 x 3 factors design.
OHP Transparency 3: 2 x 3 Factors Design
Theoretical Framework:
In the last one and a half decades the expert-novice research in various academic fieldshas discovered considerable differences in metacognitive knowledge between subjectsof higher and lower abilities. To mention just very few, in writing research BEREITER &SCARDAMALIA's (1987) distinction between so-called "knowledge telling" and"knowledge transforming" is a well-known characteristic differentiating novice and expertwriters. Despite being a well-known distinctior, it can, however, hardly obscure the factthat it only identifies differences between good and poor writers in a very general way:Novices write associatively, following the model "and then". They report, as it were,everything that comes into their mind concerning a particular topic, hence the term
3
"knowledge telling". Experts, on the other hand, apply their knowledge to create new
structures. In this way they alter existing structures or transform their knowledge as the
term "knowledge transforming" conveys.
The American research group associated with ENGLERT and RAPHAEL (ENGLERT et
al.(1989)) then investigated the metacognitive knowledge of writing in school students of
different text production abilities. Their comparison of learning disabled, low-achieving
and high-achieving school students points out clear distinctions between the three
different student groups. It is mainly the learning disabled students, who write clearly
qualitatively poorer texts, who are distinguished by less comprehensive and less
developed metacognitive knowledge. Similar differences, though to a lesser degree,exist between low-achieving and high-achieving students.
Although the research done by ENGLERT and RAPHAEL extends our understanding ofdifferent writing abilities, we still do not fully know the reason for the existing differences
in the ability to write texts. In particular, we know little about the differences which existbetween good and poor writers in a normal class concerning the knowledge how the
complex problem of writing a text can be overcome. We can also say little about the sortof metacognitive knowledge adults have at their disposal and whether differences existbetween adults who are professionally engaged in writing and those who are not. In ouropinion, analyses, just like the one presented here, are necessary for the development ofintervention programs - and later for their application in schools.
A text production model is also part of the theoretical framework of our analysis. We willnow briefly explain this model in order to give an idea which parts of the task have to becarried out and co-ordinated with each other in the course of a text production process.
OHP Transparency 4.0: Orchestra Model of Text Production
This model consists of four main components, a test component as well as an executivecomponent.
The "executive" component co-ordinates the main and test components in such a waythat it activates the components in an appropriate way. It has a wide range of knowledgeconcerning the ideal text production process and the text product characteristics at itsdisposal and makes use of this knowledge as a guiding idealized model for the co-ordination of the components. The components carry out various parts of the functions inthe text production process. They are aware of the results of the activities of other
components and can request to be active from the executive component and, if thepossibility exists, to intervene with its own function in the continuing process. The
components ai-e, however, continually called on by the executive component or, as the
case may be, activated through this. The activities of the components do not take place in
the chron-logical order of the description presented here, but are to be seen as a
complex to and from between them. This can be established from the comparison of
available and targeted results through the executive.
OHP Transparency 4.1: Enlarged Section of "Task Analysis" and"Construction of the Semantic Deep Structure"
The component "task analysis" analyses the text production task with regards to the
pregiven requirements and conditions to produce the text. It specifies the text production
task concerning its intentions of effect, its addressee reference, its text organization as
well as its rhetorical-stylistic characteristics and determines a detector model or text
anticipation which schematically anticipates the text which is to be produced (SELZ1913, 1922; DUNCKER, 1935).
The component "construction of the semantic deep structure" constructs the deep
structure which semantically constitutes the text. It does this by recalling suitable
semantic elements (idea generation) on the basis of the text anticipation from internal
(memory) and external stored knowledge (illustrations, books, media etc.). It then links
these elements with one another with the aid of appropriate relation concepts. The result
is a hierarchically organized network of semantic set-in-relations which are composed ofmicro and macro structures and the superstructure which combine these structures and
represents the "message" of the text.
OHP Transparency 4.2: Enlarged Section "Language Coding/Formation ofthe Surface Structure" and "External Representa-tion"
The component "Language Coding/Formulation of the Surface Structure" has twofunctions. On the one hand, it determines the order of thoughts for the linear surfacestructure of the text which appears in sequences. This we call imposing a chronology. Onthe other hand, it generates the correct, rhetorical-stylistic suitable, (etc.) syntax of thetext. This subfunction we call syntactising. The components, in other words, determinehow the hierarchical network of semantic set-in-relations, which are formed through
4 5
constructive processes, can be brought into a linear sequence and generate thelinguistic surface of the text so that sentences are formed. The result of imposing achronology is a list of hierarchically ordered propositions. The result of syntactisina is thecohesive linguistic text surface.
The component "external representation" represents the results of the activity of othercomponents externally. It presents, for example, the result of the relational linking of theconstruction of the semantic deep structure in a network type of form, adheres to theresult of the imposition of chronology a hierarchical list of micro and macro propositionsor represents the result of the syntactising as a sequence of linguistic surface sentences.
OHP Transparency 4.3: Enlarged Section "Test Component"
The "test component" examines, in a variety of ways, whether conditions which havebeen attained at that point and those to be attained correspond: a) Does the textanticipation correspond to the text production task in question? b) Does the constructedsemantic deep structure conform to what the text anticipatioi :equines? c) Does thegenerated surface structure or the text really express the semantic deep structure? d)Does the graphical text representation render the surface structure correctly in a writtenform? e) Does the generated text correspond to the text anticipation and thereby theoriginal writing task? The result of investigation at this particular point is transmitted tothe executive which makes use of this information when determining the future course ofthe text production processes.
The model consists of three further components, namely the components "factclarification'', "emotional/motivational state" and "internal representation".
OHP Transparency 4.4: Enlarged Section "Cognition" and "Self-Perception"
In connection with text production, it may be necessary to clarify facts which havebecome a cognitive problem. In this situation the executive brings the text productionprocess to a halt and activates the component "fact clarification". This clarifies thecognitive structure of the misunderstanding or, as the case may be, the problematic facts.Then, the executive sets the text. production process in motion again. The "factclarification" component, strictly speaking, does not belong to the text productionprocess. It is a component which also functions outside text production processes. Italways goes into action if the cognitive structure of a thing or fact has become a problem.
The component "emotional/motivational state" becomes active similar to "fact
clarification" not only in connection with the text production process, but also when a
cognitive activity acts on affective aspects. The existing "emotional/motivational state" is
transmitted to the executive. This decides what measures have to be taken and
interrupts, for example, the text production process in order to recuperate.
OHP Transparency 4.0: Orchestra Modc' of Text Production
Finally, we consider the component "internal representation" as covering the area of
mental activity which involves the internal storage and recalling of elements and
structures of knowledge.
Just as the art of playing well together exists in an orchestra in order to interpret music in
well co-ordinated interaction between the musicians and the conductor and in the
disciplined collaboration of individual musicians, so the crucial point in the outlined text
production model exists in the sensible co-ordination of individual main components
among themselves and through the executive in order to generate the text product.
Relating to this metaphor, it is comprehensible to speak about an "orchestra model of text
production".
We now turn to the analytical part of the study.
OHP Transparency 5: Experimental Design
Experimental Design:
All in all, 36 subjects took part in this study. They are uniformly divided between 11 and
15 year old school students from the Bernese secondary school (which is attended
between the ages of 11 16) and adults. The twelve subjects of each age group were
made up of six good and six poor text writers. In the student group the group of good
writers was formed from the three best writers from each of two parallel classes. The
three poorest writers from each of these classes were combined to make up the group ofpoor writers.
High school teachers co-operated in being the adult subjects. In this group the divisioninto experts or novices depended on whether the subject was professionaily involved inwriting texts or not. Those high school teachers who taught their mother-tongue
language German formed the expert group. Novices consisted of high school teachers
who did not teach German and therefore were not professionally involved in textproduction. However, they were academically trained like the experts.
For the data collection we made use of the so-called Turku-table.
OHP Transparency 6: Experimental set up: Turku-table
This table was developed at the Institute of Education of the University of Turku inFinn land, and was used in reading research. It is a specially constructed table with aglass surface and a slightly forward tilted mirror underneath this glass surface. A videocamera is positioned in front of the mirror so that pictures of a person writing at this tablewhile sitting at the opposite end of the table's camera-side can be taken as well aspictures of what the person is writing on the sheet of paper lying on the glass surface.
Each subject worked with an additional person as a pair in the experiment. For theschool student subjects this was a fellow student who was known but didn't belong to thesubject's class. For the adult subjects the additional person originated from a higher levelclass of the school (teacher training college) in which the adult subject taught. Differentroles were assigned to the two people of each pair: the subject became the tutor, theadditional person the tutee. Each pair were given the task of writing an article about theart and personage of the well-known Swiss iron-sculptor, Berhard Luginbuhl, for aschool or daily newspaper.
OHP Transparency 5: Experimental Design
The tutor and tutee had various functions as a consequence of the allocated roles: thetutor, the actual subject, had to (as senior reporter) guide the tutee (as junior reporter)through his/her task of writing a text. The tutee, on the other hand, had to follow the tutor'sadvice and was responsible for the formulation and writing down of the surface structuresentences. By organising, guiding, monitoring and giving his/her opinion on the resultsgenerated, the tutor took charge of the metacognitive or self-regulatory functions. Thetutee, on the other hand, was responsible for various cognitive functions in the textproduction process.
A twelve minute video film about the iron-sculptor Bernhard LuginbOhl and his art wasshown in order to give the subjects common content subject matter. In addition, each pairreceived six pictures of Luginbuhl himself and of some of his best known wrrks.
8
OHP Transparency 7a: Colour Picture "Luginbuhl with Butterfly-Screw onhis Head"
OHP Transparency 7b: Colour Picture of One of Luginbuhlts iron Sculp-tures
(possibly only Transparency 7b shown)
The video film was shown to the tutor and tutee prior to their carrying out the textproduction task together. It was prepared in such a way that, although it contained clearcontent material, it did not have any structured order of sequences to make a simpleretelling of the events in the film possible. In the joint session of tutor and tutee the videofilm was no longer available. However, the tutor and tutee could look at the six pictures ofLuginbOhl and his work again at any time.
OHP Transparency 8: Chronological Arrangement of the Experiment
In the first session the tutor himself wrote an informative text about LuginbOhl and hiswork. In the second session with the tutee he could therefore, corresponding to hisfunction, concentrate on the metacognitive aspects of the text production process andguide the tutee through the text production process. The video film about LuginbOhl andhis art was also shown to the tutee before the session with the tutor. The tutee, however,had no opportunity to write his/her own text before the session together. He/She wastherefore encouraged to follow the advice of the tutor in the session with him/her.
We name the process described here as data collection based on diadic instruction. Theso-called tutor doesn't solve the problem him/herself, but instructs the tutee step by stepwith the aim of enabling him/her to solve the problem. In order to do this, the tutor has tofall back on his self-regulatory knowledge of the text production process, as well ashis/her own ideas on the characteristics of text products. It is of interest to note whatadvice of a process and product related type the tutor gives the tutee or how the tutorguides the tutee through the text production task.
It was possible, with the technical apparatus described, to tape on video the problemsolving dialogue between the tutor and tutee at the same time as the notes and text werewritten down by the tutee. The session of a tutor/tutee pair lasted up to 90 minutes.
8
OHP Transparency 9: System of Categories
The 36 video tapes were transcribed. Then the statements of the tutor in each transcript
were analysed into propositions. Each proposition was assigned one of 32 categories of
a system of categories which was developed on the basis of our theoretical concepts and
the empirical data collected.
Results:
Diachronical Comparison
On the whole, adult and student sul:jects significantly differ from each other in more than
20 out of the 27 relevant categories of our category system, for example concerning (1)
the analysis of the writing task, (2) the construction of the semantic deep structure of the
text, (3) the imposing of a chronology on the constructed semantic deep structure, (4) the
test operations for evaluating the quality of the results of different writing subprocesses,
and (5) the different kinds of external representations of the results of these
subprocesses. Other differences concern (6) the way in which tutors organize their
writing process, and (7) whether they take emotional and motivational aspects intoconsideration during the writing process.
Correlation Analysis: Age and Categories
a) Ability: Good WritersAge: Grade 5, Grade 9, Adults
OHP Transparency 10: Correlation Analysis: Age and CategoriesSubjects: Good Writers of Grade 5, Grade 9, Adults
b) Ability: Poor Writers
Age: Grade 5, Grade 9, Adults
OHP Transparency 11: Correlation Analysis: Age and CategoriesSubjects: Poor Writers of Grade 5, Grade 9, Adults
The two student subject groups are altogether far more similar . They differ from eachother mainly in the attention that is paid to (1) the formulation, (2) the revision of surface-sentences, and (3) the revision of written text representation, as well as in relation to a
I9
few aspects of (4) the task analysis and (5) the linking (set in relation) of collected
knowledge elements.
Correlation Analysis: Age and Categories
c) Ability: Good Writers
Age: Grade 5, Grade 9
OHP Transparency 12: Correlation Analysis: Age and CategoriesSubjects: Good Writers of Grade 5, Grade 9
d) Ability: Poor Writers
Ability: Grade 5, Grade 9
OHP Transparency 13: Correlation Analysis: Age and CategoriesSubjects: Poor writers of Grade 5, Grade 9
Synchronical Comparison
Novice and expert adult tutors differ from each other (1) in whether they have the tuteeconsider the characteristics of the writer's audience, (2) in the kinds of test operation thetutee has to execute, (3) in the hints the tutors gives concerning the formulation of
surface-level sentences, and (4) in the tutor's hints concerning the symbolic-iconicrepresentation of generated knowledge elements.
Correlation Analysis: Ability and Categories
a) Ability: Poor and Good WritersAge: Adults
OHP Transparency 14: Correlation Analysis: Ability and CategoriesSubjects: Poor and Good Writers of Adult Subjects
Among student subjects, novice and expert tutors differ from each other in their hint.,about (1) external sources of information for generating ideas, (2) the hints aboutformulating surface-level sentences, (3) the revision of surface-level sentences, and (4)the revision of the written text representation, as well as (5) the attention they pay to the
10 11
determination of the purpose of the text to be written. However, grade 9 students differsurprisingly little from the subjects of grade 5.
Correlation Analysis: Ability and Categories
b) Ability:
Age:
Poor and Good Writers
Grade 9
OHP Transparency 15: Correlation Analysis: Ability and CategoriesSubjects: Poor and Good Writers of Grade 9
c) Ability:
Age:Poor and Good Writers
Grade 5
OHP Transparency 16: Correlation Analysis: Ability and CategoriesSubjects: Poor and Good Writers of Grade 5
Conclusion:
Our study aimed at a better understanding of the competences and deficits of writers ofdifferent ages and abilities. The most dominant factor for differences among the subjectswas the age factor. However, the observed differences are predominantly due to theadults subjects only. Much less important is the ability factor.
On the whole, the subjects from grade 5 and grade 9 are generally far more similar thanexpected. Grade 9 subjects who have almost finished school do not seem to possess awell-developed metacognitive knowledge base for writing compared to that of the adultsubjects. It seems as if students of grade 5 do not gain an elaborated metacognitiveknowledge base for writing while moving toward the end of their years at school (grade9). On the other hand, we also expected more differences between the experts andnovices among the adult subjects.
The more we understand the metacognitive knowledge base that writers rely on whiletackling writing tasks the better we will be able to design intervention programs which areintended to improve students' problem solving ability in writing. It was the aim of thisstudy to shed more light on this knowledge. The comparison of the results of the studentsubjects with those of the adult subjects gives an idea about where students' strengthsand weaknesses are in coping with writing tasks.
11 12
Reference
BEREITER, C. & SCARDAMALIA, M (1987) The Psychology of Written Composition.
Hillsdale, N.J.: Erlbaum.
DUNCKER, C (1935) Zur Psychologie des produktiven Denkens. Berlin: Springer(1974).
ENGLERT, C.S., RAPHAEL, T.E., FEAR, K.L. & ANDERSON, L.M. (1988) Students'
Metacognitive Knowledge About How to Write Informational Texts. LearningDisability Quarterly,1 1, 18-46.
SELZ, 0. (1913) Ober die Gesetze des geordneten Denkverlauff Stuttgart: Spemann.SELZ, 0. (1922) Zur Psychologie des produktiven Denkens and des lrrtums. Bonn:
Cohen.
For more detailled results write to:
Matthias BAER e-mail: [email protected] Padagogische Psychologie (APP) fax: 0041 31 + 65 37 73University of Berne
Muesmattstrasse 27
CH-3012 Berne, Switzerland
1312
How Do Expert and Novice Writers
Differ in Their Knowledge of the
\Writing Process and Its Regulation
(Metacognition) From Each Other,
and What Are the Differences in
Metacognitive Knowledge Between
Writers of Different Ages?
Matthias BAER, Armin HOLLENSTEIN, MargarethaHOFSTETTER, Michael FUCHS & Monika REBER-WYSS
Department of Pedagogical Psychology,University of Berne, Switzerland
Research supported by the Swiss National Science Foundation, Grant # 11-27673.89
Transparency 014
The Three Parts of the Project
1 to further develop a constructive structuralmodel of writing
2 to analyse the production process relatedcognitive and metacognitive abilities of writingwhich school students of different ages andadults have and
3 on the basis of 1 and 2 to assist the ability towrite texts with the aid of a longer termcognitive and metacognitive interventionprogramme
Transparency 115
Formulation of the Two Questions forAnalysis
First Focus: Young vs. Old(diachonical Comparison)
How do younger and older writersdiffer in their self-regulatory knowledge ofthe writing process (metacognition)
and in their representation of thecharacteristics of text products (textorganization)?
Second Focus: Novice vs. Expert(Synchronical Comparison)
How do good and poor writersdiffer in their self-regulatory knowledge ofthe text production process
and in their representation of thecharacteristics of text products?
16Transnnrpncni
Diagram with 2 x 3 Factors Design
A q e Ability
Expert writers Novice writers
11 year old
regular schoolstudents
grade 5
15 year old
regular schoolstudents
grade 9
adults
Transparency 3 17
EXECUTIVE
A TN H
A EL
N TG A
SK
purpose
audience
genre,kind of text,text organi-sation
linguisticaspects (style,rhetorics)
generatingknowledgeelements
relatinggeneratedknowledgeelements
micro levelset in relation
- macro levelset in relation
CONSTRUCTINGTHESEMANTIC DEEPSTRUCTURE
revising theconstructedsemantic deepstructure- missing
knowledgeelementsor relationscontradic-tions withinthe semanticdeepstructuresuperfluousknowledgeelements orrelations
Transparency 4.1Task Analysis & Construction of the Semantic Deep Structure
18
EXECUTIVE
imposing achronologyon thesemanticdeepstructure
formula-ting sur-face levelsentences(buildingspokentext)
revising thesur-face levelsentences(cohesion)
LANGUAGE
C0D
NG
Transparency 4.2 1 9Language Coding & External Representation of Products
E
symbolic/ Xiconicrepresen-tation ofknowledgeelements
networkrepresen-tation of thesemanticdeepstructure
hierarchi-cal list re-presentationof the se-mantic deepstructureproposi-tions
writtentext repre-sentation
AL
REPRESENTAT
0
EVALUATING
the proper match of
writing text anticiption semantic deep text surface final draftassignment and semantic structure and structure and and textand text deep structure text surface linguistic anticipationanticipation structure aspects
I I I I I
Transparency 4.3Test Component
EXECUTIVE
20
SEL
F
ERCEPTI
0N
emotion
motivation
C0GN
I
TI
0N
generatingknowledgeelements
=141.111
relatingknowledgeelements
21Transparency 4.4Self Perception & Cognition
EXECUTIVE
Experimental Design:
Tutors®
(,::::;)Tutees
Tutors0
high achievingwriters
average achievingwriters
low achievingwriters
11-year old regularschool students
grade 5
6 pairs
re
6 pairs0 C) C=)
15-year old regularschool students
grade 9
6 pairs
-N. e6 pairs :'') c=)
TutorsFaculty members(prof. writers) *)
TuteesTeacher studen.3(kindergarden)
TutorsFaculty members
(not prof. writers) ")
adults
6 pairs
c
N 6 pairs
(=>") Teaching: ") Teaching: Piano,Composition and Dance, Art (Painting),Literature Gymnastics, Sports,
Biology, Chemistry
22
Transparency 5
Experimental setup: Turku-table
AudioTutee
Tutor
Transparentdesktop
23Transparency 6
r
.1
rY.r
,'
....
....-
1.-4
,'
ifktt
,4p,
k-''',
)" \
'';'
erN
:t,/i
ti
'r.
4 ,,
,:
tti
f:14
1
Is!
...40
7A16
*1;0
,441
"1 t
11.,P
.,skt
efiZ
tavi
p.,7
,111
1,PA
T'4
3
'414
101.
ie
-
,.....
.:, :,
ii,,
,t
:,:.
.- ,
'.,.
....,.
....
,' y
.li..\
.,st*
,a1;
,'...)
,t,',4
). ,i
vey
....:
c ,,,
,.
., ,
' 1.(
tt'..'
,'v
t.,:i.
3,ii
.oe
f.; V
IOP4
1.,
'
1:..,
,, ,,f
,,f.c
i<1,
.. j,,
:,/',t
./;..`
'111
".:y
o?;:c
:
;.:11
:4.4
!,4,
...',/
..i.v
io,,t
,ixi
a. r
i:
',!..,
,.
...1
.,
.,
.,. p
in.0
,'''
" 1`
..77'
4k
''. t'
itt".
s't
i1
l'.//'
' Ok
''
....e
.9"
4111
'.i/.
'1.4
%.'(
,11.
4V
o 51
4 1
I,,,
,,,y
1,,..
..
iri J; 1
1.1(
ti.
.T
..(IP
,at
.. ,..
.,
..,t
.,,(.
...,..
.,
,.v.
4
t
ansp
areY
cy7a
.641
-1,1
4.1
4404
0.11
1.11
r.
t;c
r.,
0,14
1005
1,
VJ'
it01
11,1
/:,V
,1ok
ittel
'i1.
,k
.I,
'r
01?
'
!F''
BE
ST
CO
PY
AV
AIL
AB
LE40
1,, ih, a,-1
inSo
....
f ,'1:114. ieke , Z
4/1 .. 1401:10.1',,,NA
TI.bleiPihr
..-e3.-07. ...tq 44, . ,I.
No
A ,,,th'4 W
ri 1i" P i ,i.r-,,A, :-."-JC
":.` :.tia.,,;,..c.,...v,,'...44't 4%;:,tg.e,t,,,,A
V,:t:IN
11: triA".4:,,i,:y.er;g4".;-1,Z
411SivP ''
13.,,,,,.."...,..,,;:.0:;0. ,.,,, ,
.N
1ri... ''14.
.1.....
;../ , ., N,
4 ....:.i
y..
lI
,,,
V,...,...1100.1.4.,,
it - it 44 4 ...r9 ,.;,
..,.:.- ,
.,...
..
.,.; ,.....,, .22.
.,2i,..
, .2.
. ,".
...
21 .
, .1,,..
,,..- .
,.....
...,
....,.,,,...
.., ,,.
,.,........
.4,, ; ( It'fr a,,,
e n0401,: ,T
he ,_:
rt.,'" 'r,.0,....
,_ ,,10, ...?..1,
'..
.'., r. fr, a ' ...'-., rt.t.'"-k
.041. - kkt V
,'
'r ' /'( ?' I, S 4+
' --. s 's ..t1'pKtiti
16'&11..irV
):k
:10"-'1.'"ei r.t. l'..1"..:VI ''' 4 '''''.'..;t- .:17 ..14,rlit''00r.'
4 ;Attriti, 41i; .ilti %
PI .e14.4 i S.,di.t.4.z. 4,;?:.,.., N
)1..1%
, ..111.
.i,k:',:..
::. ).'1 4'.0. I.: )c .....: rt: 3 x
/1....%.o..,.411.1%
,14,.... itr.P jr stN
., 11.,,Y
10M
I{
11`;; 01;;,;1;.1.:.
! t' ......,-A,';ir.i.t.,A
'r;a1 II:14?.V Ili,.
11'1414))11101''1:'7 (.16>
kr 1. itr
iikdi 41;w
4 ii.it)'.1-ileti'4i. "Iiiij 4111"r"t er I iv. I 0 .. .
04. t..i
,i
A 41-...11
'v %,,to
I. rA",
,
tij,4
11
wrl
...;)"
itdr.'
N ', e
:4tii,
.
...,et..
1e.y
4s.41
4.t...A
.P 4+
t#Ikti
A...,ay'.
iti
At,
,(I 4! -
A.
it7-1111,i...17,::::Art.-.1". ....27.:
....7,16.kt.'s'i A
r2L`,..W
ArA
..
V. A
4!..4 I
04
Vt
0 Vt
.? I 7. %.
th.a,
:PA
4.1"41(titlh.44,./.. WA
S° I An. .
.
.77
4.1)11Id
Iry
gm
i
fitltA
'?..0.
lb4,t
t.41
S..r
L in.,
411,
ar.:t
+.
.1:.
Vlik
tieJO
.
93L
.I
s
..
''
,A t041%
..
.. A
.. , :...
-!'"e,
kvc....>tie4
. 410, , .1
'1!1r '
1)0:4;1,0 07V
f.1)XP
g
)
It
fib s/..1"Tilliy.rierV
7'tr.ictth(.0
:4
.14.4erti.. 1.10"
1'1
"".
r.
4.4 ,I /W
,VI
.1NO
)tr. 4.
..
.!IF
.I
I. .
.:1'1,L
..1,,rcA
U..1.,11J
!!:;ht.. tV
.S.
'
,
v..'
1.'
.7.,:
.4,4%
.j.500IiiiA
:(,12
;.,.Y
t
1"
Wi;i:11-41'.',
"
..t'..;.r;11
.'411i.;&°e
%,
.1'
.:,
..t
..'
'. ..
1,,
,,,,..priA
! 04,
...ir
-....:#16...''..
." '''''.:"
'c,,,,
...,
..
";.
...; .
..,, -,-.,e..: ;,.'4,,,,,..
,,t11
'.
...,
.....
.,,,o:"4, ..3.);; ...;.!,1 :.!: ltr.11,;;4. liit'i.3tP
''
'.
.'.
,-L
14,01..,ort ..1,
.tr.
,,;..
.' st
..
VIC
.A
- .r1,"
cC
'..4rt
.o.
,....0,
V
.:..4
..,--,
,,..,,
''c -.
1'711;1. .1;i
'11-' ,4.."
4.IN; rve
,
.).?
r.JO
.0
4'
,.
';
0,r
1.
411.1: f't.1 r
-
t.
,I
/441-
.1..t':1
r4.014,114,
tO).
..11.
`,
Chronological Arrangement
Time Tutor Tutee
1 . Session 1 Presentation ofVideo-tape andPictures of the Iron-Sculptor LuginbOhland His Work
2 Writing anInformative TextAbout LuginbOhland His Work
2. Session 1 Presentation ofVideo-tape andPictures of theIron-SculptorLuginbOhl and HisWork
2 Advising the Tutee Writing an
About How to Write Informative Text
an Informative Text About LuginbOhland His Work withthe Advice of theTutor
23
Transparency 8
Sal
t-pe
rcep
tions
,fe
elin
gs,
mot
ivat
ion
Org
anis
ing
the
writ
ing
proc
ess
29
]nsp
aren
cy9
Sub
ject
s' m
etac
ogni
tive
prop
ositi
ons
refe
rto
:
Tas
k an
alys
is(s
ettin
g up
text
antic
ipat
ion/
dete
ctor
mod
el)
Pur
poso
Aud
ienc
e
Gen
re, k
ind
of te
xt, t
ext
orga
nisa
tion
Ling
uist
icas
pocL
s(s
tyle
,rh
olor
Ics)
Col
lect
ing
know
ledg
eel
emen
ts
Sou
rces
of
info
rmat
ion
Inte
rnal
sour
ces
(mem
ory)
exte
rnal
sour
ces
(phe
nom
enon
,bo
oks,
med
ia
Pro
cess
of
writ
ing
Con
stru
ctio
nof
the
sem
antic
dee
pst
ruct
ure
ofth
ete
xt
Gen
erat
ing
idea
s
Set
inre
latio
n:R
elat
ing,
linki
ngco
llect
edkn
owle
dge
elem
ents
Mic
ro-le
vel
sot I
n re
latio
n(lo
cal
cohe
renc
e)
Mac
rolo
vol
sot i
n re
latio
n(g
loba
lco
here
nce)
BE
ST
CO
PY
AV
AIL
AB
LE
Rev
isin
g th
ese
man
tic d
eep
stru
ctur
e
Mis
sing
know
ledg
eel
emen
ts o
rm
issi
ngre
latio
ns
Inco
here
nce
(con
trad
ictio
nsw
ithin
the
sem
antic
dee
pst
ruct
ure)
Sup
erflu
ous
know
ledg
eel
emen
ts o
rre
latio
ns(s
triv
ing
for
sirin
g on
ce)
Pro
duct
,ex
tern
alre
pres
en-
tatio
ns
Lang
uage
codi
ng/
For
mat
ion
ofth
e su
rfac
est
ruct
ure
ofth
ete
xt
Impo
sing
ach
rono
logy
on
the
sem
antic
deep
str
uctu
re(-
Fin
ding
apa
in th
roug
hth
e co
nstr
ucte
dse
man
tic d
eep
stru
ctur
e-)
Syn
tact
isin
g:F
orm
ulat
ing
surla
co-lo
vel
sent
ence
s,(b
uild
ing
spok
en te
xt)
Rev
isin
g th
esu
rfac
e-le
vel
sert
ence
s(c
ohes
ion)
Eva
luat
ion
ofpr
oper
mat
chbe
twee
n ...
Writ
ing
assi
gnm
ent
and
text
antic
ipat
ion
Tex
tan
ticip
atio
nan
d se
man
ticde
ep s
truc
ture
Sem
antic
dee
pst
ruct
ure
and
text
sur
face
stru
ctur
e
Tex
t sur
laco
stru
ctur
e an
dch
osen
lingu
istic
aspe
cts
Fin
al (
kali
and
loxi
antic
ipat
ion
30
sym
bolic
/Ic
onic
repr
esen
ta-
tion
ofkn
owle
dge
elem
ents
not w
ork
repr
esen
tatio
nof
the
sem
antic
deep
str
uctu
re
hier
arch
ical
list
repr
esen
ta-
tion
of th
ese
man
tic d
eep
stru
ctur
epr
opos
ition
s
Writ
ten
text
ropr
oson
lalio
n
Rev
isin
g th
ew
ritte
nte
xtre
pro
sont
a-lio
n
Sub
ject
s' m
etac
ogni
tive
prop
ositi
ons
refe
rto
:
\x\\\
\\\z\
\..\
\-\\;
\\\.:
\S
ell-
\ ,\pe
rcep
tions
,fe
elin
gs,
mot
ivat
ion
\\ x\ .\
\\\\\\
\\\\\\
\\\\w
Lege
nd:
Sig
nific
ant
posi
tive
cotr
elat
ion
sign
ilica
nt n
egat
ive
corr
elat
ion
\\\\\\
\\\\\\
\\\\\\
\\\\
\\ ne
arly
sig
nilic
ant p
ositi
ve to
r:.
\s'7
..\\..
\\\.\\
\.\\.\
\*\\.
\:.\\\
..\\\.
\\:\\\
*.\\
.41/
/////.
/./.4
4/7/
.////7
/././.
/74
/7ne
arly
sig
nific
ant n
egat
ive
eon.
4///
//////
//////
//////
////
Cor
rela
tion
Ana
lysi
s A
gean
d C
ateg
orie
s
Abi
lity:
goo
d w
riter
s
Ago
. gra
de 5
grad
e 9
adul
ts
Cor
rela
tion
of A
go a
ndC
ateg
orie
s (p
oor
writ
ers
not
incl
uded
)
31n
,n
Org
aniz
ing
the
writ
ing
proc
ess
Tas
k an
alys
is(s
ettin
g up
text
antic
ipat
ion/
dete
ctor
mod
el)
Aud
ienc
e
Col
lect
ing
know
ledg
eel
emen
ts
Con
stru
ctio
nof
the
sem
antic
dee
pst
ruct
ure
ofth
ete
xt
\\\ \\
\\\\\N
"G
ener
atin
gm
illio
n\
Idea
s\\\
\\\\\\
\\st
Sou
rcos
of
info
r
Set
inre
latio
n:R
elat
ing,
linki
ngco
llect
edkn
owle
dge
elem
ents
Rev
isin
g th
ese
man
tic d
eep
Str
uctu
re
Lang
uage
codi
ng/
For
mat
ion
of
the
surf
ace
stru
ctur
eof
the
text
inte
rnal
Sou
rces
(mem
ory)
Mic
role
vel
set I
n re
latio
n(lo
cal
cohe
renc
e)
Mis
sing
know
ledg
eel
emen
ts o
r
mis
sing
rota
tions
Impo
sing
ach
rono
logy
on th
ese
man
ticde
epst
ruct
ure
(*F
indi
ng a
path
thro
ugh
the
cons
truc
ted
sem
antic
deep
stru
ctur
e')
Writ
ing
assi
gnm
ent
and
text
antic
ipat
ion
Tex
tan
ticip
atio
nan
d
sem
antic
deep
stru
ctur
e
Gen
re, k
ind
of te
xt, l
est
orga
nisa
tion
Inco
here
nce
(con
trad
ictio
nsw
ithin
the
sem
antic
dee
pst
ruct
ure)
Sem
antic
clix
epst
ruct
ure
and
text
surf
ace
stru
ctur
e
exte
rnal
Sou
rces
(phe
nom
enon
,bo
oks,
med
ia,..
)
Mac
roLe
vel s
ot in
rota
tion
(glo
bal
cohe
renc
e)
Ling
uist
icas
poct
s(s
tyle
,rh
elor
lcs)
BE
ST
CO
PY
AV
AIL
AB
LE
,\\\\N
\\\\-
`Sup
erflu
ous
\
\know
ledg
e or
.\\-\
rela
tions
\--
*,(s
triv
ing
for
\` s
irIng
ence
)X
\\\*\
\\\\\\
\
Syn
lact
isin
g:F
orm
ulat
ing
eurf
ace
leve
lse
nten
ces,
(bui
ldin
gsp
oken
text
).
Tex
tsu
rfac
est
ruct
ure
and
chos
en
lingu
istic
aspe
cts
Rev
isin
g th
esu
rfac
elov
else
nten
ces
(coh
esio
n)
Fin
al d
raft
and
text
antic
ipat
ion
32
Pro
duct
,ex
tern
alre
pre
sen-
tatio
ns
sym
bolic
/ic
onic
repr
esen
talie
n of
know
ledg
eel
emen
ts
net w
ork
repr
esen
tatio
nof
the
sem
antic
deep
str
uctu
re
\\\N
\hi
erar
chic
allis
tss
z:\
\ rep
rese
nta
\\ b
on o
f the
\se
man
tic
`\
sc".
stru
ctur
epr
opos
ition
sX
\ \\ \
Writ
ten
text
repr
esen
talio
n
Rev
isin
g th
ew
ritte
nte
xt
repr
osen
talio
n
Sub
ject
s' m
etac
ogni
tive
prop
ositi
ons
refe
rto
:
Sel
f-pe
rcep
tions
,fe
elin
gs,
mot
ivat
ion
Lege
nd:
Pro
cess
of
writ
ing
sign
ilica
nt p
ositi
ve c
orre
latio
n
sign
ifica
nt n
egat
ive
corr
elat
ion
ww
ww
N\
near
ly s
igni
fican
t pos
itive
\\\. \
\\\\\\
*\\\.
\\ .\\
\\\\.
\\.(
71/ 7
./////
, 4 //
/////:
/////7
.4 /.
/./ne
atly
sig
nific
ant n
egat
rve
4/*
//////
//////
//////
//////
///r
Cor
roia
tion
Ana
lysi
s A
gean
d C
ateg
orie
s
Abi
lity:
poo
r w
riter
s
Ago
. gra
de 5
grad
e 9
adul
ts
Cor
rela
tion
of A
go a
nd;a
tego
nes
(goo
d w
riter
s no
tin
clud
ed)
nsoa
renc
v 11
Org
aniz
ing
the
writ
ing
proc
ess
3'
Tas
k an
alys
is(s
ettin
g up
text
an ti
cipa
lion
/de
tect
orm
odel
)
Con
stru
ctio
nof
the
sem
antic
dee
pst
ruct
ure
ofth
ete
xt
Lang
uage
codi
ng/
For
mat
ion
ofth
e su
rfac
est
ruct
ure
ofth
e te
xt
Eva
luat
ion
ofpr
oper
mat
chbe
twee
n ...
Col
lect
ing
know
ledg
eel
emen
ts
Sot
inre
latio
n:R
elat
ing,
linki
ngco
llect
edkn
owle
dge
elem
ents
Rev
isin
g th
ese
man
tic d
eep
Str
uctu
re
Sou
rces
of
info
rmat
ion
Inte
rnal
sour
ces
(mem
ory)
Mic
ro.
leve
l set
inre
latio
n(lo
cal
cohe
renc
e)
Mis
sing
know
ledg
eel
emen
ts o
rm
issi
ngre
latio
ns
Impo
sing
ach
rono
logy
on
the
sem
antic
deep
str
uctu
re('F
indi
ng a
path
thro
ugh
the
cons
truc
ted
sem
antic
dee
pst
ruct
ure'
)
Writ
ing
assi
gnm
ent
and
text
antic
ipat
ion \\.n
Tex
tan
ticip
atio
nan
d
\se
man
ticde
ep\
\ str
uctu
re \
\\\\\.
\\\
Gen
re. k
ind
of te
xt. t
ext
orga
nisa
tion
exte
rnal
sources
(phe
nom
enon
,bo
oks,
med
ia...
)
Mac
role
vel s
et in
rela
tion
(glo
bal
cohe
renc
e)
\ \ \
\ \ \
\ \ \
\ \\ N;
(con
trad
iclio
ns w
ithin
the
sem
antic
deep
stru
ctur
e)\
\\\\\\
\\\\.\
\
Syn
tact
icin
g:F
orm
ulat
ing
surf
ace
leve
lso
ntor
cos,
(bui
ldin
gsp
oken
text
)
Sem
antic
deep
stru
ctur
ean
d te
xtsu
rfac
est
ruct
ure
Ling
uist
icas
pect
s(s
tyle
,rh
o to
rtes
)
\\\\\\
\\\\:
Sup
erflu
ous
\kn
owle
dge
elem
ents
or
rela
tions
(str
ivin
gfo
rst
ringo
nco)
\\\\\\
\\.\\\
\
Hav
iSin
g th
esu
rfac
elev
else
nten
ces
(coh
esio
n)
Tex
tsu
rfac
est
ruct
ure
and
chos
en
lingu
istic
aspe
cts
34
Fin
al d
roll
and
text
antic
ipat
ion
Pro
duct
,ex
tern
alre
pres
en-
tatio
ns
\\.\\\
\\\\\\
W\ s
ymbo
lic/
N:
icon
icre
pres
enta
-tio
n of
know
ledg
eel
emen
ts\
\\\\\\
*\.\\
\
net w
ork
repr
esen
tatio
nof
the
sem
antic
deep
str
uctu
re
hier
arch
ical
list
repr
osen
talio
n of
the
sem
antic
dee
pst
ruct
ure
prop
ositi
ons
Writ
ten
lext
repr
osen
talio
n
Rev
isin
g th
ew
ritte
nte
xtre
pres
enia
lion
_J
Sel
f-pe
rcep
tions
,fe
elin
gs,
mot
ivat
ion
Lege
nd:
269n
:110
n/it
posi
tive
corr
elat
ion
$lor
obco
ol n
egat
ive
corr
elat
ion
\\N\\\
\\\\\\
\\\\\\
\\\\\\
near
ly is
ond.
canl
pO
stliv
e C
orr
...\\.
\\*\.\
\.\\\.
\\*\.\
-\\\\
*\.\\
\\\\.\
\.
neal
.y S
igni
licai
neg
ativ
e C
ot'
"::e
//////
//////
//////
//////
////
Cor
rela
tion
Ana
lysi
s A
gean
d C
ateg
orie
s
Abi
lity.
goo
d w
riter
s
Age
. gra
oe 5
grad
e 9
Cor
rela
tion
01 A
ge a
nd;a
tego
nes
(poo
r w
riter
s no
tin
clud
ed)
Org
aniz
ing
the
writ
ing
proc
ess
35
Sub
ject
s' m
etac
ogni
tive
prop
ositi
ons
refe
rto
:
Tas
k an
alys
is(s
ettin
g up
text
antic
ipat
ion/
dete
ctor
mod
el)
Gen
re. k
ind
of le
st, t
ext
orga
nisa
tion
Ling
uist
icas
pect
s(s
tyle
.rn
olo
tics)
Col
lect
ing
know
iedg
eel
emen
ts
Sou
rces
of
info
rmat
ion
Into
rna
lso
urce
s(m
emor
y)
//////
//.ex
tern
also
urce
s/;
(pho
nom
en-7
/bo
oks.
////
/////.
1//
Pro
cess
of
writ
ing
Con
stru
ctio
nof
the
sem
antic
dee
pst
ruct
ure
ofth
ete
xt
Gen
erat
ing
idea
s
Set
inre
latio
n:R
elat
ing,
linki
ngco
llect
edkn
owle
dge
elem
ents
Mic
ro d
ove!
set i
n re
latio
n(ic
cal
cohe
renc
e)
Mac
ro -
leve
lse
t in
rela
tion
(glo
bal
coho
renC
e)
BE
ST
CO
PY
AV
AIL
AB
LE
Rev
isin
g th
ese
man
tic d
eep
stru
ctur
e
Mis
sing
know
ledg
eel
emen
ts o
rm
issi
ngre
latio
ns
Inco
here
nce
(con
trad
ictio
nsw
hith
In th
ese
man
tic d
eep
stru
ctur
e)
Sup
erflu
ous
know
ledg
eel
emen
ts o
rro
tatio
ns(s
triv
ing
for
strin
genc
e)
Lang
uage
codi
ng/
For
mat
ion
ofth
e su
rfac
est
ruct
ure
ofth
ete
xt
Impo
sing
ach
rono
logy
on
the
sem
antic
deep
str
uctu
re('F
indi
ng a
Rat
h th
roug
hth
e co
nstr
ucte
dse
man
tic d
eep
s tr
c lu
re")
Syn
tact
istn
g:;.
For
mul
atin
gsu
rfac
e-le
vel
sent
ence
s,(b
uild
ing
spok
en lo
x°
Rev
isin
g th
esu
rfac
ele
vet
sent
ence
s(c
ohes
ion)
Eva
luat
ion
ofpr
oper
mat
chbe
twee
n ...
Writ
ing
assi
gnm
ent
and
text
antic
ipat
ion
\\\\
\\T
ext
\ ant
icip
atio
n\
and
\\ s
eman
tic \
deep
:"'s
stru
ctur
e
Sem
antic
dee
pst
ruct
ure
and
text
sur
face
stru
ctur
e
\\.\\*
.\\\.\
\\*,
\T
ext
\\S
urfa
ce\N
and
s t r
uot
\\
\\lin
guis
tic:.\
\
s\s.
aspe
cts
N\\\
\\:\\
Fin
al d
raft
anC
text
antic
ipat
ion
3C
Pro
duct
,ex
tern
alre
pres
en-
tatio
ns
sym
bolic
/ic
onic
repr
esen
ta-
tion
ofkn
owle
dge
elem
ents
001
wor
kre
pro
sonl
atio
nof
the
sem
antic
deep
str
uctu
re
hior
arch
;cal
list
repr
esen
taLi
on o
f the
sem
antic
dee
pst
ruct
ure
prop
ositi
ons
Writ
ten
text
repr
esen
ta;Io
n
Rev
isin
g th
ew
ritte
nte
xtro
pres
enta
lion
Iran
snar
Rnr
y 1R
Lege
nd:
sign
illea
repo
litiv
e co
nela
tlen
sign
itica
nl n
eg n
ye te
net/W
on
,,\\\\
\\\\\N
\ \\\\
\\\N
\\\\\
near
ty s
igni
fican
t pos
itive
non
,\\\
.\\\
.N.\\
,/////
//////
///, (
//////
//////
/ nea
tly e
igni
lienn
l neg
ativ
e co
n///
//////
//////
//////
//////
X
Cor
rela
tion
Ana
lysi
s A
gean
d C
ateg
orie
s
Abi
lity:
poo
r w
riter
s
Age
. gra
de 5
grad
e 9
Cor
rela
tion
of A
go a
ndC
ateg
orie
s (g
ood
writ
ers
not
incl
uded
)
Org
aniz
ing
the
writ
ing
proc
ess 37
Sub
ject
s' m
etac
ogni
tive
prop
ositi
ons
refe
rto
:
Pro
cess
of
writ
ing
Tas
k an
alys
is(s
ettin
g up
text
antic
ipat
ion/
dete
ctor
mod
el)
Con
stru
ctio
nof
the
sem
antic
dee
pst
ruct
ure
ofth
ete
xt
Col
lect
ing
know
ledg
eel
emen
ts
Sot
inre
latio
n:R
elat
ing,
linki
ngco
llect
edkn
owle
dge
elem
ents
Rev
isin
g th
ose
man
tic d
eep
stru
ctur
el'u
rpos
e
Aud
ienc
e
Sou
rces
01
info
rmat
ion
Gen
erat
ing
Idea
s
Mis
sing
know
ledg
eel
omen
ts o
rm
issi
ngre
latio
ns
\\\X
\\\N
\ Mic
ro.
\\le
vel s
et In
ss
rela
tion
(loca
lco
here
nce)
\\\
X\\\
\\\\\\
\\\::
Inte
rnal
30=
05m
o ry
)(m
emor
y)G
enre
. kin
dof
text
, tex
tor
gani
satio
n
Inco
here
nce
(con
trad
ictio
nsw
hIlh
In th
ese
man
tic d
eep
stru
ctur
e)
Mac
rolo
vel
sot I
n re
latio
n(g
loba
lco
here
nce)
exte
rnal
sour
ces
(phe
nom
enon
,bo
oks,
med
ia)
Ling
uist
icas
pect
s(s
tyle
,rh
olor
ics)
Sup
orflu
Ous
know
ledg
eel
emen
ts o
rre
latio
ns(s
triv
ing
for
atrin
genc
e)
Lang
uage
codi
ng/
For
mat
ion
ofth
e su
rfac
est
ruct
ure
ofth
ete
xt
Impo
sing
ach
rono
logy
on
the
sem
antic
deep
str
uctu
re('F
Indi
ng a
path
thro
ugh
the
cons
truc
ted
sem
antic
:Jee
pst
ruct
ure"
)
Syn
tact
isin
g:F
orm
ulat
ing
surf
acel
eval
sent
ence
s,(b
uild
ing
spok
en te
xt)
Rev
isin
g th
esu
rlace
leve
lse
nten
ces
(coh
esio
n)
38
Eva
luat
ion
ofpr
oper
mat
chbe
twee
n ...
Writ
ing
aosi
gnm
ent
and
WI
antic
ipat
ion
Tex
tan
ticip
atio
nan
d se
man
ticde
ep s
truc
ture
Sem
antic
dee
pst
ruct
ure
and
text
sur
face
Str
uctu
re
Tex
t sur
taco
Stil
lClu
ro a
ndch
osen
lingu
istic
aspe
cts
Fin
al d
rat(
and
text
antic
ipat
ion
Pro
duct
,ex
tern
alre
pres
en-
tatio
ns
sym
bolic
/ic
onic
repr
uSen
talio
n of
know
ledg
eel
emen
ts
not w
ork
repr
esen
tatio
n01
the
sem
antic
deep
str
uctu
re
hier
arch
ical
list
repr
esen
talio
n of
the
sem
antic
dee
pst
ruct
ure
prop
ositi
ons
Writ
ten
lox;
rO p
rose
nla
lion
Rev
isin
g th
ew
ritte
nte
xt
repr
esen
ta-
tion
Sub
ject
s' m
etac
ognI
tive
prop
ositi
ons
refe
r to
:
//////
////1
7////
/////
% /S
elf-
// p
erce
ptio
ns,
/ /,fe
elin
gs,
//
mot
ivat
ion
/ /
Org
aniz
ing
the
writ
ing
proc
ess
Lege
nd.
sign
ifica
nt p
ositi
ve c
orre
latio
n
Sig
nific
ant n
ap 'A
live
corr
elat
ion
\\\N
\\\-
\\\N
\\\\\\
\\N\\
near
ly s
igni
fican
t pos
itive
cor
r...\
\\ ...
\\\\.\
\ .\\
.\\ .\
\N.\\
s
/1/
//.//.
/////
/1/ n
early
sig
nilic
ant n
egat
ive
corr
//441
////,%
&%
%/7
/2X
XX
X.4
61,X
XO
Cor
rela
tion
Ana
lysi
sA
bilit
y an
d C
ateg
orie
s
Abi
lity:
poo
r an
dgo
od w
riter
s
Age
: adu
lts
Cor
rela
tion
01A
bilit
y an
dC
ateg
orie
s (g
rade
5 a
ndgr
ade
9 no
t inc
lude
d)
39ns
par
e nc
y14
Tas
k an
alys
is(s
ettin
g up
text
antic
ipat
ion/
dete
ctor
mod
el)
Pur
pose
Oon
ro, k
ind
of te
xt, t
ext
orga
nisa
tion
Ling
uist
icas
pect
s(s
tyle
,M
oiet
ies)
Col
lect
ing
know
ledg
eel
emen
ts
Sou
rces
of
info
rmat
ion
Inte
rnal
sour
ces
(mem
ory)
exte
rnal
sour
ces
(phe
nom
enon
,bo
oks,
med
ia...
)
Pro
cess
of
writ
ing
Con
stru
ctio
nof
the
sem
antic
dee
pst
ruct
ure
ofth
ete
xt
Gen
erat
ing
idea
s
Set
inre
latio
n:R
elat
ing,
linki
ngco
llect
edkn
owle
dge
elem
ents
Mic
ro lo
ve
Ise
t In
rota
tion
(loca
lco
here
nce)
Mac
roov
else
t In
rota
tion
(glo
bal
cohe
renc
e)
BE
ST
CO
PY
AV
AIL
AB
LE
Rev
isin
g th
ese
man
tic d
eep
stru
ctur
e
Mis
sing
know
ledg
eel
emen
ts o
rm
issi
ngre
latio
ns
inco
here
nce
(con
trad
ictio
nsw
hIth
in th
ose
man
tic d
eep
stru
ctur
e)
Sup
erflu
ous
know
ledg
eel
emen
ts o
rre
latio
ns(s
triv
ing
lot
strin
gonc
e)
Lang
uage
codi
ng/
For
mat
ion
ofth
e su
rfac
est
ruct
ure
ofth
ete
xt
n
Impo
sing
ach
rono
logy
on
the
sem
antic
doop
str
uctu
re('F
indi
ng a
path
thro
ugh
the
cons
truc
ted
sem
antic
dee
pst
ruct
ure"
)
%7/
//// S
ynla
ctis
ing:
For
mul
atin
g,
surf
ace
--//
leve
l/ s
ente
nces
,(b
uild
ing
/ spo
ken
text
) /
//////
////.
Rev
isin
g ,h
esu
rfac
elev
else
nten
ce!:
(coh
esio
n)
40
Eva
luat
ion
ofpr
oper
mat
chbe
twee
n ...
Will
ing
assi
gnm
ent
and
text
antic
ipat
ion
Tex
tan
ticip
atio
nan
d se
man
ticde
ep s
truc
ture
Sem
antic
deep
Str
uctu
rean
d te
xtsu
rfac
est
ruct
ure
Tex
tsu
rfac
est
ruct
ure
and
chos
enlin
guis
ticas
pect
s
Fin
al d
raft
and
loxl
antic
ipat
ion
Pro
duct
,ex
tern
alre
pres
en-
tatio
ns
sym
bolic
/ic
onic
repr
osen
taeo
n of
know
ledg
eel
emen
ts
not w
ork
repr
esen
tatio
nof
the
sem
antic
deep
str
uctu
re
hie
rarc
hica
list
repr
osen
ta-
I--
lion
of th
ese
man
tic d
eep
stru
ctur
epr
opos
ition
s
Writ
ten
text
repr
esen
talion
Rev
isin
g th
ew
ritte
nte
xtre
pres
onla
lion
Sel
f -pe
rcep
tions
,fe
elin
gs,
mot
ivat
ion
Legend:
sign
ifica
nt n
egat
ive
corr
elat
ion
\\\\\\
\\\\\\
\N\\\
\\\N
\\\\\
\N n
early
sio
nil c
ant p
ositi
ve c
orr
:\ .\\
\.\\N
\\: \\
Ns.
\\.\\
.\\*
..\\\
;/////
//////
//////
//////
//////
,ne
arly
sig
nific
ant n
egat
ive
con.
;,///
//////
//////
//////
//////
Cor
rela
tion
Ana
lysi
sA
bilit
y an
d C
ateg
orie
s
Abi
lity:
poo
r an
dgo
od w
riter
s
Age
. gra
de 9
Cor
rela
tion
ofA
bilit
y an
dC
ateg
orie
s (g
rade
5 a
ndad
ults
not
incl
uded
)
ansp
aren
cy 1
5
Sub
ject
s' m
etac
ogni
tive
prop
ositi
ons
refe
rto
:
Pro
cess
of
writ
ing
Org
aniz
ing
the
writ
ing
proc
ess
41
Tas
k an
alys
is(s
ettin
g up
text
antic
ipat
ion/
dete
ctor
mod
el)
Pur
pose
Con
stru
ctio
nof
ILO
sem
antic
dee
pst
ruct
ure
ofth
ete
xt
Col
lect
ing
know
ledg
eel
emen
ts
Aud
ienc
e
Sou
rces
of
in fo
r m
a lio
n
Gen
re, k
ind
of te
xt. t
ext
orga
nisa
tion
Inte
rnal
sour
ces
(mem
ory)
Ling
uist
icas
pect
s(s
tyle
,rh
olor
ics)
exte
rnal
sour
ces
(phe
nom
eno
n, b
ooks
,m
odla
...)
Gen
oiat
ing
idea
s
Set
inro
la li
on:
Rel
atin
g,lin
king
colle
cted
know
ledg
eel
emen
ts
Mic
role
vel
sot I
n re
latio
n(lo
cal
cohe
renc
e)
Mac
role
vel
set I
n re
latio
n(g
loba
lco
here
nce)
RE
ST
CM
AV
AIL
AB
LE
Rev
isin
g th
ese
man
tic d
eep
stru
ctur
e
Mis
sing
know
ledg
eel
emen
ts o
rm
issi
ngre
latio
ns
Inco
here
nce
(con
Ira
d ic
lions
with
inth
e se
man
ticde
ep
stru
ctur
e)
Sup
erflu
ous
know
ledg
eel
emen
ts o
rro
tatio
ns(s
triv
ing
for
sirin
genc
e)
Lang
uage
codi
ng/
For
mat
ion
ofth
e su
rfac
est
ruct
ure
ofth
ete
xt
impo
sing
ach
rono
logy
on
the
sem
antic
deep
str
uctu
re("
Fin
ding
apa
th th
roug
hth
e co
nstr
ucte
dse
man
tic d
eep
stru
ctur
e')
Syn
lact
isin
g:F
orm
ulat
ing
surf
acee
vel
sent
ence
s,(b
uild
ing
spok
en te
xt)
Rev
isin
g th
esu
rfac
edev
else
nten
ces
(coh
esio
n)
Eva
luat
ion
ofpr
oper
mat
chbe
twee
n ...
Writ
ing
assi
gnm
ent
and
text
antic
ipat
ion
Tex
tan
ticip
atio
nan
d se
man
ticde
ep s
truc
ture
Sem
antic
dee
pst
ruct
ure
and
text
sur
face
stru
ctur
e
Tex
t sur
face
stru
ctur
e an
dch
osen
lingu
istic
aspe
cts
Fin
al d
raft
and
text
antic
ipat
ion
42
Pro
duct
,ex
tern
alre
pres
en-
tatio
ns
sym
bolic
/ic
onic
repr
osen
talio
n of
know
ledg
eel
emen
ts
not w
ork
repr
esen
tatio
nof
the
sem
antic
deep
str
uctu
re
hier
arch
ical
list
repr
osen
tato
n of
the
sem
antic
dee
pst
ruct
ure
prop
ositi
ons
Writ
ten
text
repr
osen
talio
n
Rev
isin
g th
ew
ritte
nte
xtre
pro
sent
slio
n
Sub
ject
s' m
etac
ogni
tive
prop
ositi
ons
refe
rto
:
Sel
f-pe
rcep
tions
,fe
elin
gs,
Mot
ivat
ion
Org
aniz
ing
the
writ
ing
proc
ess
Lege
nd:
eign
hcan
l pos
mve
cor
reia
lion
men
dica
nt n
egat
ive
corr
elat
ion
.\\\\\
\\\\\\
\\.\\\
-\\\\
\\\\\
near
ly S
igni
fican
t poS
ttNe
Cot
tk.
\\N\\.
\\N
\\\\ .
\\\\ \
N. \
\\\r/
//////
//////
://///
//////
/.,e
;ne
arly
ign4
ican
t neg
ativ
e 55
t.///
//////
//////
/////
7n.
Cor
rela
tion
Ana
lysi
sA
bilit
y an
d C
ateg
orie
s
Abi
lity.
poo
r an
dgo
od w
riter
s
Age
. gra
de 5
Cor
rela
tion
ofA
bilit
y an
dC
ateg
orie
s (g
rade
9 a
ndad
ults
not
incl
uded
)
4)
)nsp
aren
cy 1
6
Tas
k an
alys
is(s
ellin
g up
text
antic
ipat
ion/
dete
ctor
mod
el)
Aud
ienc
e
Gen
re, k
ind
of te
xt, t
ext
orga
nisa
tion
t--
Ling
uist
icas
pect
s(s
tyle
,rh
utor
ics;
Col
lect
ing
know
ledg
eel
emen
ts
Sou
rces
of
info
rmat
ion
inte
rnal
sour
ces
(mem
ory)
exte
rnal
sour
ces
(phe
nom
enon
,bo
oks,
med
ia...
)
Pro
cess
of
writ
ing
Con
stru
ctio
nof
the
sem
antic
dee
pst
ruct
ure
ofth
ete
xt
Gen
erat
ing
idea
s
Set
inre
latio
n:R
elat
ing,
linki
ngco
llect
edkn
owle
dge
elem
ents
MIc
rolo
vel
set i
n re
latio
n(lo
cal
cohe
renc
e)
Mac
role
vel
set i
n re
latio
n(g
loba
lco
here
nce)
BE
ST
CO
PY
AV
AIL
AB
LE
Rev
isin
g th
ese
man
tic d
eep
stru
ctur
e
Mis
sing
know
ledg
eel
emen
ts o
rm
issi
ngre
latio
ns
Inco
here
nce
(con
trad
ictio
nsw
ithin
the
sem
antic
dee
pst
ruct
ure)
Sup
erflu
ous
know
ledg
eel
emen
ts o
rre
latio
ns(s
triv
ing
for
strI
ngen
ce)
Lang
uage
codi
ng/
For
mat
ion
ofth
e su
rfac
est
ruct
ure
ofth
ete
xt
Impo
sing
ach
rono
logy
on
the
sem
antic
deep
str
uctu
re("
Fin
ding
apa
th th
roug
hth
e co
nstr
ucte
dse
man
tic d
eep
stru
ctur
e')
X/
////
Syn
lacl
Isin
g:F
orm
ulat
ing
surf
ace-
leve
lse
nten
ces,
(bui
ldin
g/ s
poke
n te
xt)
4///
//////
///
Rev
isin
g th
eS
urfa
ce-
leve
lse
nten
ces
(coh
esio
n)
Eva
luat
ion
ofpr
oper
mat
chbe
twee
n ...
Writ
ing
assi
gnm
ent
and
text
antic
ipat
ion
Tex
tan
ticip
atio
nan
d se
man
ticde
ep s
truc
ture
]Sem
antic
dee
pst
ruct
ure
and
text
sur
face
stru
ctur
e
44
Tex
t sur
face
stru
ctur
e an
dch
osen
lingu
istic
aspe
cts
Fin
al d
raft
and
text
antic
ipat
ion
if
Pro
duct
,ex
tern
alre
pres
en-
tatio
ns
sym
bolic
/ic
onic
repr
esen
talio
n 01
know
ledg
eel
emen
ts
not w
orn
repr
esen
tatio
nof
the
sem
antic
deep
str
uctu
re
hier
arch
ical
list
repr
o se
nta
tion
of th
ese
man
tic d
eep
stru
ctur
epr
opos
ition
s
Writ
ten
text
repr
esen
talio
n
Rev
isin
g th
ew
ritte
nte
xtre
pres
enta
lion