ED 257 096

DORMS? RESUME

CS 208, 929

AMOR otalL, Sauli; Vahipasii, Andeli'TITLE. . Internstional Study of WrAtteh Composition.PUB DATE 85 .. .

=TS ' 26.; Papervreminad at the.AnnUalikeeting 'of the'International Writing Convention (Norwich, England,March 31-April 44,1965).Figure 4 'may not be legibledue ,to small .print: « ,

PUB TYPE e Reports - Descriptive A141) --.Speeches/Conference'e*lepers (150) ______ .-

EDRS,PRICB ; Mf0 l/PC01 Plus Postage.DESCRIPTORS Elementary Secondary Education; Foreign Countries;

*Grading; *Nolistit Evaluation;`*liternational. Xducational Exchange; Writing (Composktion); *Writing. .

.

, Evaluation; Writing.PrOcesses; *Writing'ResearchIDENTIFIERS *Writingirasks/

.

44BSTRACT.

, , . . Fifteen member 'Countries of the International- , Association for the Evaluation of Educational Achievement IIIM) ° *

collaborated on a project that had four aims; (1) to contribute to\-, the conceptualization of, ;the domain of.wrfoting, (2) to develop an

internationally appropriate sot of writing tasks and a system forassessing,compositions,13rto,deicribe-recent developments and thecurrent state of instruction in.written4omObsition, and .4) to.identify factor( that explain Jiff rakes and patterns rd the

4\,Mperformance of written- composition. A set of writtngtnsks was.presented.in each country to °never more of the followlOg populationgroups: (1)-the end of Primefreducatioi (Population A), (2) the endof compulsory (*cation (Population B), and 13) the prtuniversity

.

year (Population C).Alhe recommended nalple 'gizmo were 50 classes forPopulations A.and C, and 100 classes for Population B. Problemsrelating to the construction of writing.taski, the allocation,ofwriting tasks, and the scoring of student scripts ;tops addressed. Themost appropriate scoring system combined a holistiC, ' 0

overall - impression; arking and ah analytical marking.. It wasconcluded that in spite of a great efforethe scoring metric was tooe lastic to allow robust cross-national comparisons of scores on any ..

one task or group, of tasks. A sample scoring sheet, tables; anddiagrams are included: (DI') .

S.

********r*************************************************************** Reproductions supplied by EDRS are the best that can be made* from the original document. i

***********************************c**********************************

a.1

.,...y-i ... us.oarquirmosiOP IOUCATION - A. NATIONAL INSTITUTE OF EDUCATION '.., -;

. EDUCATIONAL R!SOURCES INFORMATION -.=v.

X .

CENTER (ERIC)

This ducurnent has been reproduced INpeened limn the person or organdation

originating it. r fI' 1

f i Minor changes have been made to Inspects* ...1reprodu5tion.qualitY ,

INTERNATIONAL STUDY OF WRITTEN COMPOSITION

. Sauli Tgkeila & Anneii Vah4passiInstitute for Educational ResearchI, University of Jyvasky14

INTRODUCTION

- - - - - - _

Points of view or opinions stated in this cloaca. '",.1

mint do not necessarily represent official NIEpation or policy.

VERMISSION TO REPRODUCE THISMATERIAL HAS BEEN GRANTED BY

O j5Saul/ Takala

Anneliyahipassi

; TO THE EDU9ATIdNAL RESOURCES.INFORMATION CENTER (ERIC)."

In. this paper we will give a brief account of an'.42-,-going

,international 'study of written,, composition.' Fifteeli .member

countries-.

f the IEA (The International Asiociation for .the

Evaluation of. Educational Achievement) from all over. thd world

-arecoilaborating on the project. A fairly, large set of writing'

tasks has. been presented in some countries. to one population of

students, in some to two. and in some to all three populations,

representipg end of primary education (Pali. A), end of compulsory

edccation (Pop B), and the,pre-university year.(Pbp'C), respecti-.

'velY/ (for: a detailed account, see Table 1, Appendix 1). The

123 recommended minimum sample sizes were 50 classes for Populations'co25. A and C and:'100 for POO B.

417C.

- We will .first place the study within its larger context.

g: After that we proceed to describe the aims of the project, its

8 desixgn and scope..

. We will.give particular emphasis to the task. I.

t3.

and to the scoring procedures. 'We, will also present some resules.OD

concerning the pilot stages t the project.

V)V

We wish to point out that this, is not the "official"

or complete presentation of the IEA Written Composition Study. We'4

Vhave chosen to highlight some aspects of the study and have added

some personal reflections and interpretations.

. Paper psented at an International Writing Convention, Schoolof Education, University of-East Anglia, Norwich, March 31-April4, 1955.

", ' ki :17:-

b

CONTEXT OFAHE STUDY

Since the late 1950's, a numbgc of "educational researcher

and research institutions have been workiqg on an empirically

'oriented 5ompanative research' program. A small feasgbIlitY'study

was carried outlay the IEA in 1959-1962. Thii was followed by a

First Mathematits Study in the mid 1 O's. (with 12 countres'

school system ..partitipating) 'and .by ii x studies in 1970-1971.

These covered Science J19 coUntries;, Reading Comprehensi9n

Liter4ture (10), French as a foreign language '(8), English as a.I

:

fore gn.language*.(10Y, and Civit Education .(19). On-gbing studies. .

co prise 'a second mathematics and science study and a study .9f-

cftssrdom activities. Since 1979, work has been carried out on an

. )international study of written composition.

-

i

A number of international and national reports have been r

published onthe.Completed studies. Similarly, a great number of0

journal articles have been published, as a bitaliography of IEA-

related publications shows.

MANAGEMENT OE THE STUDY

Like all LEA studies, the Written Compositions study has a

..---complex management of ucture. The International Project Coumcil

at its annual meeting makes general policy decisions. Dr. Alan C.

Purves from the University of Iltinois at Urbana-Champaign is the

Chair of the IPC. More specific planning is the responsibility of

the International Steering Committee, chaired by Anneli VAhApassi

from Finland. Members are Tom Gorman (England & Wales) Judit

K!adar-Fulop (Hungary) , Eva Baker (USA), Alan C. Purves (USA),

1Htldo Wesdorp tNetherldnds, Until 1995) Pai Obanya (Nigeria) and

Vaimo Konttinen (Finland). Sauti Takala (Finland) is the Interna-

tional Coordinator and Elaine Degenhart (USA) Deputy Coordinator.

3 BEST COPY AVAILABLE

oda

Q.

a

.(7.;;;4;. t q, ..

From January 1981 to'kefail 1984, the International 'Coordinating

Center was located At .the Curriculum Laboratory.(UIUC)...Sipie the

fall of 1984the.coordinating of the project is managed doi tly

6

by the Curriculum. Laboratory and-the Institute for Educational.

. I.. 4 . '..

Research, University of Jyvaikyla. 4.

. .

National Research CoordinatOr constitute the International

'Study. Committee, ; which has .met., a few times to discuss the. ,

,implementation the study according to common plans.t 4

The costs of international coordination have been paid by,

the IEA, while4the national costs of implementing the study Are

paid. by each particiOatina country..

1

41

D,SISNL AND AIMS ©F THE STUDY

The IEA International Study of Written Composition seeks to

accomplish the following tasks:'

(1) to 'contribute to the conceptuallizaton of the domain "ofwriting and partitularly the domain of school-based written-composition,

(2) 'to develop a an internationally appropriate set of writing4, tasks and a system fOr assessinq compositions Whith is

applicable across countries/icKool sytems and acrosslanguag'es,-

(3) to describe regent developments and the,current state ofinstruction in written composition in the participatingcountries/schooLesystems, and

11

(4) to identify factors which explain differences and patternsin the performahce of written composition' and otheroutcomes, with particular attention to cultural bac-ground,curriculum and teaching practices. .

The independent and de endept variables of the studyand their

prr.iumed relationships are illustrated in Figure ,l. We will

Place Figure 1 about here.

BEST COPY AVAILABLE41

16

:4:

.

7.'d.1.fs".-;7 ,... r a

begih. with th conceptualization andioperationalization of ther

dependent variable. .

I.

DEPENDENT VAtRIABLE: .WRITING TASKS AND SCORES

. _:.,.'

Problems relateg to tbe deRendeot .T ;';'':". --,--

The greatest challenge in the study has been the work on the %

,:.

dependent variable. Here we can give only a brief, accoUnt'of it ..-.._

..(

. .

.A -

::::.(for- a detailed Account,- se Vahapasst,. 1911?, 1983). Three'

.

problem areasareas had to be addressed: ,.:

;

.....,J... Pcoblems related to:the'constructiOn of writing tasks.... 2.,..,

......---------\ (a) What is the total domain of writing,, especially school- .°. N

..., ,

.

I based -domain of written composition? This. required .". - ,,-

I conceptual analysis and synthesis.. s..

,-i.

.

(b) What is the appropriate sample from the total domain'for-the students-concerned?--What sub-domains should be.

I J .-'

included in the set of writing tasks?

_;4:t

(c) What is?he appropriate systems for specifying thetasks?

11.-Problems related to the allocation of writing tasks

(d) How should the actual writing tatks be formulated?

, ,..

(e) What kind of tasks should be presented to each, of thethree student populations? .

( . t .

,*..,

1

(-0 How should populations tie linked_thrQugh common tasks?..,

..

(g) How many tasks should/can each student be asked torespond to?

(i) If task rotation is necessary, how should it be done.so as to maximize the 'information obtained andminimize problems that are.related to rotation?

III. Problems related to rating of student scripts1

(j) What rating system should be used (e.g., holistic"primary trait, analytial)?

.10

(1::) How con a rating system be constructed which, can beapplied in a comparable way in all papiicipating'countries?

It is pcIssibletto focus tiere on a few of these problems only.

\-BEST COPY AVAILABLE4 .

..c.'-..7.:-;; ''PT:.--:...',44''',. .;:irl:--v: .... -IF,. ....i4 -v i.,,...,4. -...? . , ; :=t- ,f,,-3,-e , -; , .. 3 ".A.44f'e.rf,0,-; '' .ii.::.'fr e' '

. ) 4E1. I

V .'; .sr

i.

;a c

:,. I°, I

-gaeskiksatkgq'ai tftt Romaka ef lici.tits-

I.

0

y

After ark extensive review of the problem area nd after

4having tried several systems and discarded' them, Ithe .following

approach (Figure 2) was adopted (for a more.detailed ccount.; see .%.

Vahapassi, 982, 19831..( .. ,

.N." J .

I

===================ft==m=Plaice Fkgure 2 here_

========================4

i.

r

,-,.

,. .

.

,....

. The system relates the primary purpose and audience 434 writihao..

. ,,,

..

.

the . level and object of cognitive procese& iryeialved. in the ..:,;4

,.

.

,

L .0

';10..

writing task. This gives us a.typology and we can test it in.sOme

respects hy .trying to place various commun ication tasks 1 and

writing products in its cells.

1

This has been illustrated;

--......,.

4';:,!..:-,;,

'..:;,

,,: .-

J.r:.

...,.'

say that up to now, the model has serve' as a very useful tool in .,,.,...

.-r

solving a number4 problem. We will return to the model later ,.

.......

on... ,,...,

I°selectively in Figure 2. Without-going.into detail suffice it to

ft.1

gampling of tasks from .the 4omgin

In the IEA study, writing has been considered 'an' act of :

communication about some topic, betweerNhe writer and the

reader-(s) (addressee/audience). The writer has a certain purpose

in writing..and his ultimate goal is communicative effectiveness

(effe=tive expression).\ Typically the writer.producqs a. text

alone, without the immediate feedback from (cooperation/negotia-.

Lion with) the reader.!-This latter.-aspect seems to be one of the

most important features that distiGguishes writing from face-to-

face conversational interaction (Takali, 4982, 1983). The above

kscussion has been sholdWn schematically in Figure 3.

== = = = =

Place Figure 3 here%0 =========================

6 BEST COPY AVAILABLE,:.,",,,?.; :4.

k.t......ili..,':,:t.,W...- .."1-.... ,t.,,,A,, .`-4 -,' ., .Q-,11,511

0 ' , .. 4U'!

a40 I A

t , 1

. re ........do .

The tasks of thi IEA WMItten. Composdtion Stydy' can bes .

rAe4ted to' .1.12- domain specification grid (Figure,2) and to 'the' ..

rhetorical model of writing (Figure 3) as follows: . %:.... .. .

.. .. 1.

'11).Tasks that emphasize the perspectivy of the writer..

Task 5: JWrite -a personal. story ,

Task 8:' write a "free" composition on an ambiguous and

. .

,4evocative pictorial stimulus .

(2) Tasks that.emphasize'thepernpective of the topic .

:

1

Task 2: summarize "a teHtTask,34-yetell a story (1.n a shorter form)Task 4a: describe a ritual mask . 0

Tag. 4b: -describe a process of doing swathingTAA. 7: write a reflective essay

I

(3)..Taskssthat emphasize, the perspective.of the reader

Task 6:' try to persuade the readerto share the writer'sstrong view about *something r

(4) Task's that hae several perspectives

1

.%,.. :,

. . .ATask: la: 'describe a.desired bicycle to an uncle who,

. ,

. wishes to. buy one...as_abirthday...present.... .. :. -,

Task lb: describe oneself to &penfriend whom the.studpnt is going to visit so as to makeit possible for the penriend to identifythe student as he comes to meet the student

'1-1 ,Task°1c: write. a note to the princi3Oal/headmaste)-

cancelingsa scheduled meeting

. Task id: leave a message, at home lling where thestudent has-gone after school,

Task le: .*writa a letter applying for an advertisedsummer job

P

1 Note tnat die writer-reader social status ands the topfC is

varied systematically.

Task 9: write a letter to a younger student who iscoming to study at the same school as the

Awriter, telling the new student how he /shsshquld write in the new school-to get good.grades.

If we focus on the purpose of the tasks, we can see that

Tusks 2, 3, 4 and 9 have a predominantly informational

purpose. The expressive purpose is dominant in Tasks 5 and 8.

_ : .RFcr rilirt AV `1

ti

77

. .

,

.

Tasks is and 6 have a. per'suasive,... Opur sq!....Task 7' has an

explicat*ve/interpretive, purpose.

If we pay attention to the cognitvoi-stuctu c of the,tasks,

we might 'sugg;esfi Tasks lc,. id, -24 isold i have a :

.)structure.bAsed mainly on a.temporal organization (an account of

events). Tasks _la, lb and 4a have a strdcture based mainly on a

.

spatial organization (an account of. the physical charat4eristict

-:q.47:,:s

,of objects). Tasks le,' 6., 7, 8 and 9 have a stru444.4ri based. 1p7,1, .

.. .

.

.

.mainly on a-logical organization lan°accountlof.ideas and.thought, 1 '.

structures). Other classifications are,.offcource, possible. For.

an example of actual tasks, see Append4x.2.,

Scoring scheme

SGor pg sygleiD

After considering and .ying out various scoring., systems

'(holistic, primary trait; aria al), it was agreed that the

most appropriate one was a system which. combin0 a holistic

overall impression marking and analytical. marking. Figure 4 is

. an attempt to summarize the conceptualization work, done in deve-

loping thescoring system.

Place Figure 4 'here===r======================= O`

4.?

The overall construct "student writing" is divided into 'two

sub-- construc "writing, compence" and "writing preferences".

'We are c ce ed here mainly with Witing competence, which is

taken to consist of "discour4e7dtructuring competence" (or

rhetorical codkietence) and of "text-producing competence".

Rhietorical competence is assumed to require both "cognitive

competence" and "social competence".

r.

BES,T COPY AVAILABLE.

0

./ .:p , r z v.

,e4(4

i;

4'4

4 Cognitive competence is manifested illocutionary effecti-

veness, 1..e., j, the abilityof-wri,,teris to makeTeaders easily5,

i:

recognize the cc,1municativeinten. Tilts is .opeeationalized. in -

our scoring scheme as rated "Quality and scope of ideas" and

ap rated success in the'"Organization and-Presehtati'on of Con-

tent". In sortie tasks, writers also aim at perlocutionary effecti-.

veness, i.e., they want to get the'readers to do something. This

may mean that writers Piish to have readers entertained or 'have

them change their minds,, or even do ,something concrete after such

a ,chande of mind.. Such perlocutionary effectiveness presupposes

awarenes of readers' norms, attitudes,. etc. In our scoring scheme

this is operationalizedlin part, aJ "Style and Appropriateness of

Tone".'

tis ,

Text-producing competence presupposes guistic compe-

tence" and "motor compet4nce". These are operiatiOnalizad in our. ,

scoring scheme as rated competence in using ."grammatical fea-

tures",, 'spelling and or.tographic conventions", and "handwriting.

and neatness", all rated in accordance with national norms., For a

sample scoring seheme, see Appendix 3.

The fact that raters are readers and will, in spite of

their attempt to rate compositions as objectively as possible,

according to tp categories of the scoring scheme, have a

subjective response to the scripts. For this reason,, we have

added .a category, "Response of. Rater", which allows raters to

indicate such a response. This is another way of trying to get an

estimate of perlocutionary effectiveness.

Scoring scales

171 scoring .scheme is lupplemented.with a scoring scale,

which consists of a numberaof compositions that exemplify various

scale point values of the scheme. There were no particular

problems in having a consensus on the adopted scoring scheme. On

BEST COPY AVAILABLE;

/

...

'S

the other hand, the construction of the scoring Scales was a

difficult and .laborious task.

There were ,sveral, problems that had to be .addressed.. The

most difficult one was how to establish internationally} agreed!

scores for,compositionswrittLa in several different lang4ages..A

number of'alternative4'were considered (e.4., translation into

several languages, use of bilingual raters, having an

international rating panel). After long. discussions, it was

agreed -that the most.'444a0ible solution was as follows: (1) Pilot

test the tasks: (2) National Centers select a Set of. compositions

representing the whole performance range, and rate .thqm according

to national 'norms. (3.) Translate these proposed "benchmark" cow-4

positions carefully 1 into English, follnwinga common iset. of

g4delines. (4). AssemI31e juries with an i'nternatignal composition0

and have them rate t0

e (sometimes translated) compositions after-

the juries have been'trained.according to the established train-'

ing system. (5) In rating the compositions of the main study, use.

the original benchmarks (written in the .source language) with

their international scores in rater training and as .a reference

in final rating. (6) Check the comparability of rating in parti-

cipating countries by means of a common "international calibra-

Lion which'is a set of compositions with a high degree of

consensus.

Some Results

The system described in the above was applied for the first

time during the third International Study Committee Meeting in

Urbana in the spring of 1984.

Four international juries consisting of five National

Research. Coordinators were set up. Members of theSteering Com-

mittee served as jury leaders and native speakers o4 English as

. KTSCIPY AVAILABLE

e.

rapporteurs,

J.7p-=.

The estab1 ishment benchmarks and ,cal brati on sets was

iii muted .mainly to Population B. Altogether some 1,400

compositions had to be rated, discussed, and c ommented on: Each

Jury member scored about 500 cap sitio s. Jury members knew only

\1t the target aspect that had been given. a nationally proposed

initial score. The 'compositions were handled in the following

way:

- if there was a 4 out of 5 agreement on the nationallyproposed target aspect, the score was international validated(whether, it was identical with or lower or higher than .thenational suggestiOn)-..

r,y1

- if there.. was blower agreement or if a jury wanted another "Jury :t6. have a look at some compositions for .whateverreasons such compositions syWerq referred to restoring

- if some composi\ions received, highly -uniform stores' on al4target aspectsl'they weresingled out as candidates for .the

,international c4libriation sot.. +k-\

; -\

The results of\ttick Urbana scoring session have been pa4 lyv .

analyzed and can be briefly summarized is follows:,.

'

1

1- .i

. .

Si) Ey using a common set-Of-compositions scored by all Jury'

members after they have finished scoringa particular task, it

was foun4 that there was full agreement in 22% of all *cases,

78,2%. agreement with one scale point deviance. This waa lower

than we had hoped, but not unusual even when raters belong to the

same ihterPretive commuiity. The produced benchmark scores do not

constitute.as gooda yardsti6k as is desirable.

(2) There was some systematic difference in Jury leniency ( meana

'of all differences in means between all pairs of Juries .326). /. I /t(Z) The average standard deviations of Juries varied only /.

?slightly. 46e

(4) The perception of quality of composition* was to a larege

extent convergent between juries as median corlations ranging

between .83 and .91 on different\ aspects show.

(5) Jury homogeneity WAS fairly. good (.85 or higher, Cronbach's

E-EST an AVAILABLE

3'

N . .:;'.% ..,:..7 `..`,..,-;z2.)...5-,:;.; l'f'-'- ;do .

4alpha).

,!/.

(6) There was considerable. scale sit'inkage: the 5-point scale6 /

shrank to a.3-point *scale in- about of-the cases. This is to

f

LA.

..T.;

1: .

some. extent an artifact, because i'n some cases countrieshad a. . \. \

- .\ .;,_.

z .smalljpil9t sample and reported that they sometimes felt obliged ,

ti, ..--,f,

/

,

to propose a 5 even if they felt that that was too high.

(7Y . Marie problematic are the 10'4. of all cases /when the

. /international jury revered the scale (e.g.,. givirigZ5-4 r.3-2-

4 when the prdmciped .order of merit was 1-3-5);

The results of the .second icoring_meeting held in Frascati

late 1984 have.not been fOly-.analysed y., The implications of,the results were discussed at a eeting. I'1 '

, (3 .:.:.

r

in Hamburg ifs'-February, 1985. It was concluded that in spite of a '

I,

. great etfort' thwscoring metric/is too plastic to.allow 'robust. .::.:.

;7..

/cI

cross-national coTpariSpns/ of/ scored. on ,rly*one or group, of .

i. /

tasks. .Since there is quite a lot of interest, in slime---,--- .

... .

international companison, some further work iejheeded. There are

1 ^:1plans.,tottry to improve the comparability ysing one-task and une

,population .only. A robust international scale is created and a1,L

subsample of compositions are rescored. This'one task will then

be used for limited international Comparisons.

For anyone working in the assessment bf wrItten.cdWposition

the restAltyw an 4.nternational attempt to create a fully compa-

rable scaring procedure.are not surprising. In fact, we have been

somewhat surprised that we have been able to solve as manyit*

problems as is the casP While robust comparisons of achievement

across countries are doubtful, we do not consider that we have

failed in our attempt. It appears that tasks 1, 2, and 3 (see

Design and Aims of the Study above) will be successfully

accomplished. That is more 0-lan some df us dared to expect.

7 BEST 'COPY AVAILABLE

. 4\\, t

.-.-%

INDEPENDiONT ,VARIABLES1k

.. \ i

igdre..1 presets the major constructs that underly the

,ilarge set of variables used as iyidependent. variabies. D taired

:information h44 been gatheredby'Auestionnaires9. interviems and $

written case studies about the context of written composition

1 'teaching.1

, ,

. The gathered information (covering hundreds of ivaeiablea)

wi I be used ,mainl y to, provide detailed descriptions of the;k' re'

context and status of'writing instruction in participating coun-../.°

tries. A smaller subset of variables Will probably be used in.

some multivariate anal-ye fi hnd model testing.

We are particularly,intergs in the portraits o( writin4

instruction that the very speci TeaaclIer Questionnaire data ...,"...

% ,.. .

, ;la e

will p/

;''

.....;:y.)..

. :IC..

REPORT IN OF RES4LTS\ - . ` :.11:1$

!.: -r/'' There well be both'internaiional ird national reporting

the findings ..of,,the study. Three international -repOrts. are.\

planned. The first volume will deal wittl the dependent variable.

It will give a detailed account.of the probrems and issues in-

volved in constructing an international set of writing tasks and

in 'sc ing student scripts using an internationally mreed-upon

s:oring procedure. The report is expected to be in manuscript

form by the end of 1985.

The second volume will give a detailed description of the

context and practices of writing instruction in the participating

countries.

c The third volume will present thd main results of the study.

This will probably be a set of parallel national .portraits with

some international comparisons. The extent of comparisons will

BEST COPY MAIM

in the participating countries.

/SOME CONCLUDING REMARKS

In conclusion, we would like tip make some personal

observatio.is. The IEA writing study has been the most difficult..

(And laborious task ever undertaken by either of us. It has fore

Us to tackle Many problems without-being able to draw on any go

models. The essure of deadlines has also been veru,heavy.p

On e other. hand, the IEA study has been , more rewarding-w4o.

than y other project we have been working on.' Facing new

has been a cdnstant challenge and given us an

opportunity to learn a lot. The chance to work on our prdblems

with dedicated and experienced colleagues from all over the Work,

has been invaluable in finding justified and .acceptable

solutionq. It has. so taught us to be more sensitive to cultural.Je

similarities and d fferances. We have come to be more aware of

the variety of ways in which writing instruction can'be arranged.tr

\ On the basis of the enthusiastic reception theIEA.study has .

P-

.::.;

a'l

:

had in our country, we are convinced that the Project will have...:{

;;,,

a major impact on the teaching of compositions in Finland.,.::.!:.

I

Through our participation in the ICEA study, we in Finland will .1,,

get a 4irse1good national/assessment of 'writing. In addition to'5*

that, we will be able to' r'llate our situation to that in several

. 03,

other countries. We haim established close contacts with manyHt.....,4

..,4,

colleagues in other countries.

There are plans to store a representative sample of student,

scripts in an international student text corpus. Several signs4

indicate that there is growing interest to move from the

assessment stage tp. a stage, when we can take a close look at

the compositions themsqlves. .A number of cognitive, linguistic,

. - ,, '. / -

°

I .

,

rhetorical,, cros cultural etc studies have been tentatively.....,

4, ...,

sketched during ourirecent meetings.

REFERENCES

e

S. (1982). On the origins,' communicative parameters and

processes of writing. ggalgatkpg ih $ducstkoq& (3), 209-230.

Takala, S. T- (1983).' The domain of writing. In S. Takala lc A..

VitiAPaesi, OD the ammtifiop gl Ihs19120mili9i writing (PP.

1-61). University cf JyVAskylAi Institute 'for .Educational.. 0

,Research, Report No. 333.4 ...

"JahApassi, A. (1982). On the specification of the .domai-n---of-------t..

. .

school' writing.Ivaluatiop in EdgcationA 5(3).,165-289.

I

VAhapa,ssi, A. (1983).,

the epecification of the -domain__.-of---, .

school-writing. In S. Takala & A.. Vahapiiiii', Q. 04egr4-i

specification bf the jgmlin of wtiting (pp. .63-93). University

4

of Jyvtaskyla, Institute for Educational Reseacch, Report No. .

4333.

.rik*

A) %..- -;1I ,.-;

, 14/7 N4 ....f

, '4,4

15 .

BEST COPY AVAILABLE

:A?

.

- .

..:.;

Appendix 1

... I' .

-. .., J.

WIGS 1. Tested tacit. pee. populaacet in participating cow. Ica (plun acme. betel:growl ir4koranti on ).-. '

r

;...;';..q.:.:;;\

Awstralla

Chile

England

Federal lepublicof Germany

Finland

Military

Indones11

Italy

leteerliind6 .

Nigeria

Sweden

iaalland

United States

I

,

Grades. Aimsdledlee Class Sliming Testing Oates for Tested Populations

Populations'. Year Is Medlin

Si be Tested Hellas Age School/Grade 9ess Slee.

A I C C Al CAIIC

A I C . 12.13 1S-11 17-19

I C 13-14 17711

A t115A I C, II 14 .17-11

.% IS.I711 I -C 12 If .17

A I C

A. C

C

A I C

CD

.Y

S.IO

II

12 g 21

31

1,27

2$

S .5 12 3$ 32 37

II 31 30

3ts 13 23

18 28

1 .11 13.

(17) (010) (412)

10-12 IS-II 1722 4 II 13 SO 41

12.13 16+14 11.15 I .9 11-12

17.1$ 12 35

11-13 15-11 111-17 S II 11-12 21 25

Tasted' taiga/population

Is 16 lc Id te 2 3 la

C -

11. S I - I . 1.0 II I.

Au

VT" s

//1 Alll

8 .

-

A IIC 8

.9

,,

A A/8I SIC. .

A ,A A4 A AAAA A/8 AA A 8/C 1/C A. AnI* . -

p C C

. 5 11

4

ic.t

8 I 8

I A/8 WIC, iVe Co 1/C

8 .1 8/Cl/C. l';

A A A

PA 1/1 1/1/C 11/C .1/C ..1/C

11' # #

vs. AMC-AP 4-4/0 .4 AA- r

15 AA AA A/I Al/CS A - ,11 Ail A/11/C -

A/IA/IA/I A NC - A - - A/11/C -

C C. C C -

IS Ail A/1 A/11 A 8/C 8/C A Ail S All A/I/C lie -

V

BEST COPY AVAILABLE.

ti

".x :-;,,,,q0 'M :.r

. STUDENT

CHARACTERISTICS

is

NONE CHARACTERISTICS

s.

COMMUNITY CHARACTERISTICS

SCHOOL CHARACTERISTICS

4

TEACHER

PERCEPTIOh

Of scoot:

AND STUDENTS

CLASS

*MrZATION

alma,

CURRICULUM

AoTEACHINO

PRACTICES

-Objectives

-tasks

naterlals

timid( forvritIng -

-feedback

4d

'140k

011

STUDENT

PERCEPTIONS

STUDENT

.LEARNING

OUTCOMES

)

OpPORIIINITY

TO LEARN

TEACHER

CHARACTERISTICS

P!GURE 1% General Model of the Study (Main Constructs)

,

Tasks In Relation to the Domain of School Writing

Cognitive

DominantIntention/Purpose

ProcessingR&R000CE

..-

ORGANIZE/REORGANIZE.

INVENT /GENERATE 1

Mary-Pri on-nary ntAudience

...._....

.

Facts Ideas

Visual images, facts '. mental States,% events

Usti

Ideas, mental states,

alternative worlds.

.

To learn(meta-

filigUS1)

S

e1

f

Copying

hitting Dictation

. ,

Retell a story Resume #2t(heard or read). Summary .%1/ .

Outline(:) Paraphrasing

Comments on book margins

MetaphorsAnalogies

..

,

To conveyexotigps,feelimotivenbs()

SelfOthers

scram ofconsciousness

Personal story. Portrayil(E%%sr. 7 Personal-dfary

Persona) letter

Reflective writing.

-.PersonalThe traditional

,

To inform(referee-tial)

0t

h

ers .

.

dinstructionWOnto Fill to

a ToneMessage

(1)

Narrative report °fractions ()WS ,2% Description

TechnicalTelegram description(:)Announcement- illographyCircelv Science report/'Message (!) experiment

Expository writingf...DefinittOn

-.Academic essayarticle

--Rook review.-Commentiry

literary genres

and modes

can be placed

ender one-

or more

of these four

purposes.

To con-wince/persuade(canna.my*/

0t

hers

.

Citation fromauthority/expert

letter of Advertilementapplication Letter of to

advice '.7Stateliest of wassail (1)

views, opinions

Argumentative/persuasivewriting 0-- Editorial

...Critical essay/article (2)

To enter-tale. de-lifht.

please

(poetic)

.

OthersQuote poetryand prose

Given an ending. 1

create *story Word portraitCreate an ending or sketchRetell a story. Causerie ®

, ""y4'

Entertainmentwriting i,

--Parody 0)-- Rhymes

- PRIMARY RODEOr DISCOURSE

DOCUMENTATIVEDISCOURSE

EONSIALWE DISCOOsENarrative "--,,Descriptive

Explanatory.. . _ ....

EXPLORATORY DISCOURSEInterpretive Literary(Expository/ _-Arnoromositivo

. ./-:-

ti''.

4011, filostrooltI-

611 !Will [RISC(,

Ototo,..so.ttrocteloo

tOtiessr.o.orirloctog)COP toots

falOr114O049Coo 'roc;

Ivo. Co,. Cots tips, tripe Como Como ofImlet, Illy of of II of Iso,soltof- *tots* Tome isoyo ACti Stills. o *coswiling tag III

,(1 0

or Dotes.

. 1

%Comm,. lot lol. Totyrliol No

ism.. Ng. .

=1st% Cytosokmoo NM.

1

N

cotmg oimmemi

1adostaretsi .LION[)Isstrotttoo.

litemo.

0 egsior

Jets tot 104WAO PO :to Solllos L ol

,1

t

q

"'amt. Mee. Stool .ossoo C26110 SIM,80110 . 00 I Cow. Col tome

1 ,toot* Iloc

. (

IWOomilltl ' Ilyl IA from. Seaton, ol000rilloo Solt. mr000r Class'''. tecomis.;A ' 40000,010100 Mal MO 06 lort of tome Sm..stsototsoo ost loolorot OrstrAvosolt Pootootsof 6.14t WW1 (mtillsoli Co q.[

of loaf Nally) Cmoyoettoot Motto')Matt.' Natio') .."

1,p I

1

a

. .

4 .

Appendpc.

-

:

. ,

TASK 6 .-'4ARGU1ENTATIVE/PERSUASIVE IMPOSITION

Popul d t ion Student RaterI

-..".7':',4"..`'..713!t,1: :*t7.:.:7C '

Sample Scoring Sheet.,

I.

Overall Impression

.NAL QuOtx add Scope of Content

1. Significance of wh4 is said

4.

2." Argumentation/Exposttion

B. Organizailon and Presentation of-Content

3. Organization of the. whole text

44. Organization of iubsiunits

C. Style and Tone

6,..- , ^

WITHOESPECT TO WRITER's CHOICE %1Or.AIN AND RfADERSHIP

.`

140

I nadequa te1

.4. 44

Excellent:3- 4 S-.

4

.;

..s+f

.

OfteMMOMMIN

Po

. 111=111

' 6;

aomi; 411wommovo. .

14141

Is...._ 111Mil OMOIMMWM.

1111.111111

fMlV

=15. Choice of consistency of tone

...k,,

..

--.......... .........

, .

6. Choice of words.and phrases, ientence. .-...;

,o

structures, and larger units of

discourse.

\

\

El. Grammatical features

\

E. Spelling and orthographic.conventions

.

.......... - ........, ............ i,

4

A!.

0 ..f

\ . ..,

F. Handwriting and neatness,

........ ......... ......... .........,i..,., --...,.'

.

. ._........, 1.f.

...

.-

Nimmomormimmi

=111111

'FAL

WWW610

G. Response of Rater. . .

Low Medium High..

1 4

,

8. My interest in the composition is4momommirmm 444smorm r

9.)tiy sense of being persuaded by1

wilimmimmem, ,

the composition