Student writing development is not likely to happen without revision. In this computer-assisted learning era, revision is increasingly practiced as a self-regulated process. Students can regulate their revision behaviors through cognitive strategies like goal-setting or social strategies like peer review. We examined possible effects of writing goals together with peer review on college students' self-regulated revision in order to better understand how to enhance that revision. We asked one research question in particular. What kinds and effects of revisions did students produce in the digital learning context of the current study? 52 U.S. college students participated in this study. During the 16week semester, students were instructed to write two papers (with writing goals written at the end), each in two drafts. When they finished writing each draft, they submitted documents into an online writing and peer review system (Peerceptiv), reviewed peer papers, gave required comments and ratings, wrote reflections based on peer review, and then revised and resubmitted their drafts. We found that students made targeted revisions. High level revisions occupied a bigger percentage in terms of changes of sentence numbers. Revision-caused writing score improvement was not tremendous, but statistically significant.
Deep, lasting learning of any subject matter requires self-regulation, including
writing. Writing, as a complex cognitive task, entails self-regulation to orchestrate the
sub-processes, and adjust writing efforts for better writing achievement (Schunk &
Zimmerman, 2007; Harris, Graham, MacArthur, Reid, & Mason, 2011). The
improvement in writing is gradual, which needs serious efforts from both instructors and
students. Computer technology provides facilitation tools to increase writing practice
and feedback and helps transforming writing into a self-regulated learning process.
High levels of self-regulation are required from a student writer in the ―usually
“selfplanned, self-initiated and self-sustained” writing activities
Goal setting and peer review are both important self-regulation strategies to
enhancewriting and revision. Even though goal setting is an effective and flexible
intervention technique, there has been less focus in the literature on goal setting
interventions than on teacher and student feedback
Revision can be identified and revealed through the types of revision writers make
in writing, since revision leaves a record of what has been changed from first draft to
later drafts. Faigley and Witte (1981) tested their taxonomy of revision changes by
coding and comparing only six research samples. Sengupta (1998) further
coding of revision in terms of types, sizes and functions is more explicit in
determining the multi-facet dimensions of revision, the workable
counting of the numbers of changes and thus more reader-friendly to be
applied. Later revision categorizations are largely based on the above
mentioned studies to cover the operations, types and functions of
revisions
Specifically we asked one foundational question about self-regulated revision behaviors. The question is: What kinds and effects of revision do students produce?
The participants were 52 undergraduate students from a variety of majors enrolled in an introductory Cognitive Psychology course at this large public university in the Northeastern United States. These 52 students agreed to include their class tasks in this research study, and then completed the writing tasks and surveys analyzed here. These enrolled students are with slightly above average mean on the college entrance examination SAT scores (max = 800, national means of approximately 500): SAT math mean=606, SD=69; SAT verbal mean=611, SD=80; SAT writing mean =596, SD=75. Student writers who complete all four assignments including writing goals, peer review, self-reflection from peer review and revision fall into 52. 3
Across the 52 student draft 1/draft 2 pairings, a total of 476 revision changes were found. In terms of frequency, each student made an average of 9 revisions (keeping in mind that some revisions could affect multiple sentences). That is, students were making targeted revisions, rather than entirely rewriting their papers. In terms of size, revision has been made within, equal to or beyond a sentence. Within sentence boundary, revision happened at format, letter, word, phrase and clause level. Beyond sentence boundary, revision happened at sentential and paragraphical level. There was no case of rewriting the paper. 61% revision fell into sizes within sentence boundary, and 39% revision fell into sizes beyond sentence boundary. The categories and percentages are as follows.
citation 1% conclusion 1% other interestingness clarity and 3% 0.25% support introduction 22%
0.25% grammar 28% wording 33% image
7% flow 5%
In terms of effects, revision was explored in two ways. We measured whether or not students improved paper ratings over drafts. Then, we count frequencies of function of revisions to measure the contribution of revision. Revision was measured by the improvement of scores from first draft to second draft. A paired t-test (t(51)=3.92, p<.002, Cohen‘s d=0.54) of paper ratings from draft 1 to draft 2 showed paper ratings improved by a mean of 0.21 (out of 7). Slightly more lower-scored Draft 1 papers tend to improve more in Paper 2 round. It suggests that students did not make revoluMean # 3.4
2.5 4
tion
NA
4.0
2.3
ness
11.0
tion
6.0
To summarize, revisions were measured in terms of types, functions, paper rating
improvements and sentence numbers. It is in line with writing research tradition
Grammar 1.0
1.2