OCR Document - Lizanne Lafontaine

advertisement

The Study of Peer Response Groups on the Improvement of Writing Skills in Grade 4 Students

Living in Minority and Majority contexts

Paper presented at the

2006 AERA Annual Meeting

Friday, April 7- Tuesday, April 11, 2006

San Francisco by Sylvie Blain, Ph.D.

Professor at the Faculté des sciences de l’éducation

Université de Moncton

Moncton, New Brunswick

Canada, E1A 3E9 blains@umoncton.ca

and Lizanne Lafontaine, Ph.D.

Professor at the Département des sciences de l’éducation

Université du Québec en Outaouais

283, boulevard Alexandre-Taché

Case postale 1250, succursale Hull

Gatineau (Québec)

Canada J8X 3X7 lizanne.lafontaine@uqo.ca

This research is financed by the

Social Science and Humanities Research Council of Canada

Blain and Lafontaine (2006) The Study of Peer Response Groups… a) Objectives

The disappointing results on written productions by French-speaking students raise the issue of learning and teaching French in Canada, whether it is seen in a linguistic minority milieu as in

New-Brunswick, or in a linguistic majority context as can be found in Québec. Indeed, the writing skills of New-Brunswick pupils, as assessed by the DIEPE, group in 1995 and the Cabinet of

Education of Canada, 1999, have been shown to be poor, especially in spelling, grammar and syntax.

The students of the Outaouais (a western Québec region close to the Ontario Border) have also been assessed with scores of 70.9 % by the end of high school, these scores being inferior to the provincial average of 74.9 % on written French evaluations in that province (Ministère de l’Éducation du Québec, 2001).

This presentation will give the results of a three-year research conducted on 72 fourth-grade students, half of them living in a French majority milieu, and the other half living in French minority context. The main objective of this research was to determine whether comments made during peer response groups (PRG) were incorporated by students into their final copies. The other objective was to examine a possible link between revisions made with and without peer feedback and improvement of the written text. b) Theoretical Framework

The conceptual framework of this study is based on sociolinguistic theories of language learning, particularly Vygotsky's developmental theory that considers learning as a result of social interaction. According to Vygotsky (1978), group work provides learners with the opportunity to acquire cognitive skills that are just beyond the individual student's grasp. Peer response groups are presumed to provide a favorable setting for students to benefit from a "zone of proximal development".

Models of oral teaching and learning (Dolz and Schneuwly, in 1998; Lafontaine, 2001) suggest ways of teaching oral language, but only in very specific activities (formal discussion, debate, radio interview, etc.). Some studies effectively examined the phenomenon of construction of oral conversation and explanatory interactions during Language Arts periods (formal discussion, debate, radio interview, etc.). More recently, the scientific community considered oral language as a very important part of literacy (Germain and Netten, 2005). However, studies examining the development of oral communication between the pupils for the purpose of mutual aid and knowledge construction pertaining to writing, as is the case with PRG, are few and deserve to be intensified.

This lack of data spurs a continued search by the scientific community for significant factors which determine the success or failure of such feedback activities.

According to four studies conducted with young second-language learners (Ammon, 1985;

Blain and Painchaud, 1999; Samway, 1987; Urzùa, 1987), it seems that peer feedback has a favourable influence on young students' writing skills. Grade 3 learners improve their writing abilities when teachers let them write frequently, give them verbal feedback on content first, and put the emphasis on the communicative aspect of the activity by publishing their students’ texts (Ammon, 1985). In

Samway's (1987) study, 54% of the revisions made by 15 students from Grades 2 to 6 were influenced by verbal feedback received during writing conferences and/or peer response groups. It seems that these revisions improved the written quality of the texts. We cannot, however, speculate on the effect

1

Blain and Lafontaine (2006) The Study of Peer Response Groups… of peer feedback, as this variable was not isolated in these studies. Peer response groups helped two

Grade 4 and two Grade 6 classrooms of children revise their texts effectively as reported in a study by

Urzùa (1987). They developed a sense of audience, a sense of voice and a sense of empowerment in language. Urzùa (1987) qualified her data as modest beginnings and she raised the issue of the relationship between peer feedback and the quality of the written texts. The results of Blain and

Painchaud, (1999) indicated that not only 55% of peer comments were integrated into texts, but that the quality of the writing also improved between the first draft and the final copy for Grade 5 second language learners. Children as young as the participants in these studies are therefore capable of providing feedback in order to improve certain aspects of their friends’ texts (Blain and Painchaud,

1999).

Studies of elementary students in English or French Language Arts indicate that there are some positive effects of peer feedback. Peer response groups seemed to motivate Grade 4 to 6 writers

(Russell, 1983; Sudol et Sudol, 1991) who were able to take into account peer suggestions when they revised their copy. (Messier, 1989). These suggestions are especially useful when they are specific

(Messier, 1989). (Neubert et McNellis, 1990) suggest that is possible to teach students how to give specific feedback with adequate training. Revisions made within the context of verbal feedback resulted in an improvement in writing quality between drafts for Grade 6 students (Brakel-Olson, 1990) as well as Grade 4 students (Blain, 2001). The results of this action research, performed in the framework of learning written French in a minority situation, indicate that elementary school teachers can make use of peers during the revision phase of the writing process (Blain, 2001).

No study has verified the effects of peer feedback over a long period of time between control and experimental groups. These previous considerations and the lack of empirical data as to the effectiveness of peer feedback for young writers have led us to the following research questions:

1.

Are peer comments given during the peer response groups integrated as revisions in subsequent drafts?

2.

Did the revisions made with and without the help of peers contribute to the improvement of writing skills?

3.

Do the results differ regionally? c) Methods and Data sources

The 72 participants in this research were chosen from four fourth grade classes, two in the

Québec province, the only province in Canada where French speakers are in majority, and two in

New Brunswick, where French speakers represent 33% of the population. There were one control group and one experimental group attending French schools in each province. The experimental groups received peer feedback on every second composition. In order to perform an in-depth analysis of comments made during peer response groups, utterances were recorded for 16 participants, 8 in each province. In New-Brunswick, three out of these eight students had bilingual family backgrounds where one parent was Francophone and the other Anglophone, whereas in

Québec, only one student had such a bilingual family background. In each team of four, there were one skilled, two average and one non-skilled writers.

Neither the participants nor their teachers were familiar with the peer review process. They were trained by the researcher following procedures similar to those proposed by Hudelson (1993),

Samway (1987) and Urzùa (1987).

2

Blain and Lafontaine (2006) The Study of Peer Response Groups…

Two kinds of data were collected: 1) first draft and final copy of six compositions written by all participants once a month during a six months period; 2) the transcripts of the peer response group sessions for those 16 participants.

The data analysis consisted of three main operations. First, we compared the utterances of the 16 participants with their drafts to verify if peer comments had influenced revisions. This was accomplished by analyzing transcripts of peer response groups as well as the revisions made using an adapted version of the Blain (2001) categories.

Secondly, we verified whether revisions improved writing quality by assessing the first draft and the final copy of six compositions for all participants. Three measures were used: 1) number of words as counted by Microsoft Word® software; 2) holistic scores using the Carroll (1980) scale; 3) analytic scores obtained from three content-related criteria (communication, organization, coherence) and five form-related criteria (syntax, vocabulary, punctuation, grammar and spelling).

Two raters, who were unaware of the research objectives, were trained to grade the 852 texts.

 Inter-judge Pearson correlation varied from 0.281 to 0.938 for the holistic score and the eight analytic scores criteria. The correlation was higher for assessing surface features of the texts (ranging

0.764 to 0.938), with the exception of lexicon (only 0.281). We chose to eliminate this last criterion.

The correlation for the content text features ranged from 0.5 to 0.678 and the correlation for holistic score was 0.53. We then opted to eliminate the texts where the inter-judge gap was too important, in an attempt to improve the correlation. As a result, the inter-judge reliability for holistic scores rose to

0.59 and content text features rose from 0.573 to 0.847. For the remaining texts, we calculated the average scores between the 2 judges.

Thirdly, we compared all the results to determine if revisions, comments and progress were similar or different for students living in a majority milieu as compared to students living in a minority milieu. d) Results

First, we questioned whether peer comments made during the peer response groups were reinvested as revisions into subsequent drafts. As a rule, students from New-Brunswick revised their texts by integrating 28% of the comments made on the content and 57% of the comments on form that were made during the peer response groups. Students from Québec integrated 15% of the comments made about content and 60% of the comments made in regards to the form of their texts.

Both groups of participants in either setting tended to integrate a greater proportion of comments about the form of their texts than comments that were made about the content. Revisions of content for both groups were successful when comments were precise and repeated by more than one person of the group. Very often, for form revisions, only one suggestion was necessary for the comment to be integrated in the next revision.

Thirty-seven percent (New-Brunswick) and 45% (Québec) of the comments consisted of positive feedback and did not lead to any change. Writers in New-Brunswick did not consider peer comments for 27% of the total interactions and writers in Québec did similarly for 48% of the comments they received - an indication of the sense of authorship of the children. Also, out of the revisions that were made without any apparent peers' suggestions, some were influenced by the conversations that occurred during writing conference. See graph 1 at the end of the article.

i

3

Blain and Lafontaine (2006) The Study of Peer Response Groups…

Most suggestions for form corrections were accurate, and when they were not, writers rarely integrated them. Form corrections were more numerous for grammar (87 in Québec and 34 in New-

Brunswick) and spelling (68 in Québec and 52 in New-Brunswick). Comments made during peer response groups were also more numerous for these categories (grammar: 87 in Québec and 43 in

New-Brunswick; spelling: 67 in Québec and 62 in New-Brunswick). See graph 2 at the end of the article.

ii

In conclusion, the analysis of data for question 1 indicate that the corrections for form were more frequently integrated in subsequent texts than those that were made about the content, and this, in both provinces. Considerations of verbal comments made about content were greater in New-

Brunswick than in Québec. As form is concerned, it was almost the same in both milieus.

Secondly, in regard to improvement of text quality between first draft and final copy, no significant differences were found between control and experimental groups as the multivariate analysis seem to indicate for the number of words, for the holistic scores and for the global scores of analytic assessment. See graphs 3, 4 and 5 at the end of the article. iii  This lack of significant improvement of the writing quality might be due to the fact that the experimental group received peer-feedback for only half of the compositions. Furthermore, control groups received help to revise and to improve their texts from their teacher who taught them revision strategies. Both groups were then experimental and both conditions seemed to have positive effect on the writing quality.

Differences were significant between minority and majority settings for the percentage of form errors (grammar, syntax, punctuation and spelling, for all texts), both for experimental and control groups as we can see in graph 6. iv These results corroborated results found in studies where

French speakers living in minority milieu of Canada obtained lower scores on their writing skills. e) Educational and scientific importance of the study

This research provides data on the way peer verbal feedback helps young writers in revising their texts. It also provides data on the type of comments and revisions that lead to improvement of writing quality. From a theoretical point of view, it contributes to the growing body of evidence on teaching "writing as a process" and on the use of peer feedback for young writers. It also raises intriguing questions about the differences in learning writing skills between the majority and the minority contexts in education.

4

Blain and Lafontaine (2006) The Study of Peer Response Groups…

References

Ammon, P. (1985). Helping Children to Write in English as a Second Language: Some Observations and Some Hypotheses. In S.W. Freedman (Ed.), The Acquisition of Writing Language:

Response and Revision (pp. 65-84). Norwood, NJ: Ablex.

Blain, S. (2001). Study of Verbal Peer Feedback on the Improvement of the Quality of Writing and the Transfer of Knowledge in Francophone Student in Grade 4 Living in a Minority Situation in

Canada. Language, Culture and Curriculum . 14 (2), 156-170.

Blain, S. et Painchaud, G. (1999). L'impact de la rétroaction verbale des pairs sur l'amélioration des compositions des élèves de 5 e

année en immersion française. The Canadian Modern Language

Review / La revue canadienne des langues vivantes . 56 (1).

Brakel-Olson, V.L. (1990). The Revising Processes of Sixth-Grade Writers with and Without Peer

Feedback. The Journal of Educational Research, 84 (1), 22-29.

Carroll, B. J. (1980). Testing communicative competence.

Oxford: Pergamon Institute of English.

Dolz, J. et Schneuwly, B. (1998).

Pour un enseignement de l’oral

. Paris : ESF éditeur.

Germain, C. and Netten, J. (2005). Place et rôle de l’oral dans l’enseignement/apprentissage d’une

L2. Babylonia, 2, 7-10.

Hudelson, S. (1988). “Writing in a second language.” Annual Review of Applied Linguistics , 9 , 210-

222.

Lafontaine, L. (2001).

Élaboration d’un modèle didactique de la production orale en classe de français langue maternelle au secondaire . Thèse de doctorat inédite. Montréal : Université du

Québec à Montréal.

Messier, L. (1989). Influence de la rétroaction verbale des pairs sur la révision de productions

écrites d'écoliers de sixième année . Mémoire de maîtrise inédit, Université de Montréal,

Montréal.

Neubert, G.A. et McNellis, S.J. (1990). Peer Response: Teaching Specific Revision Suggestions.

English Journal , 79 (5), 52-56.

Russell, C. (1983). Putting research into practice: Conferencing with young writers. Language Arts,

60 (3), 333-340.

Samway, K.D. (1987). The Writing Processes of Non-Native English Speaking Children in

Elementary Grades . Unpublished Ph.D. dissertation., University of Rochester, Rochester.

Sudol, D. and Sudol, P. (1991). Another story: Putting Graves, Calking into practice and perspective.

Language Arts, 68, 292-300

Urzùa, C. (1987). "You Stop Too Soon": Second Language Children Composing and Revising.

TESOL Quarterly, 21 (2), 279-304.

Vygotsky, L.S. (1978). Pensée et langage . Éditions sociales.

5

Blain and Lafontaine (2006) The Study of Peer Response Groups… i

Graph 1 :

Percentage of Interactions Leading to No Changes, Integrated into Texts, and Ignored (N=455)

60

50

40

30

20

10

0 content QC

(n=136) content NB

(n=81) form QC

(n=145) form NB

(n=93) no changes integrated into text ignored ii

Graph 2 :

90

80

70

60

50

40

30

20

10

0

Number of Interactions and Correct and Incorrect Revisions for

Each Province Relating to Form int.QC

rev.QC

int.NB

rev.NB

syntax inc. syn.

punctuation inc. punc.

lexicon inc. lex.

grammar inc. gram.

spelling inc. spelling int. = interactions inc. = incorrect punc. = punctuation gram. = grammar rev. = revisions syn. = syntax lex. = lexicon

6

Blain and Lafontaine (2006) The Study of Peer Response Groups… iii

Graph 3 : Writing quality : number of word (Sig=0,89)

150

145

140

135

130

125

120

145

146

130

133 first draft final copy control experimental

Graph 4

:

Writing quality : holistic score (Sig=0,961)

3,3

3,25

3,2

3,15

3,1

3,05

3

2,95

3,133

3,252

3,077

3,191 first draft final copy control experimental

Graph 5 : Writing quality : total analytic score (Sig= 0,248)

19

19

18,5

18

17,5

17

16,5

16

15,5

15

14,5

17

16

16,6 control experimental first draft final copy

7

Blain and Lafontaine (2006) The Study of Peer Response Groups… iv

Graph 6 : Writing quality : differences between contexts (Sig=0,00 ; % of error)

12

11,94 11,81 11,81

10

8

6

8,56

9,21

7,58

6,45

QC

NB

3,98

4

2

0 punctuation grammar spelling syntax

8

Download