When multi-peers give better advice than an expert: The type and impact of feedback given by students and an expert on student writing

This study was to investigate the role of reciprocal peer reviews in improving writing quality. To serve the goal, the sources of feedback were manipulated as a subject-matter expert, a single peer, and multiple peers. Of major interest was the hypothesis that feedback from multiple peers would be more effective than that by an expert in improving novices' writing quality.; Student participants were matched based on their writing skills and randomly assigned to either a single expert feedback (SE) condition, a single peer feedback (SP) condition, or a multiple peer feedback (MP) condition. Novice writers in SE received feedback from the expert, those in SP received feedback from a single peer, and those in MP received feedback from six peers. The core procedure was that writers turned in their first drafts, reviewers reviewed papers, writers received feedback based on their feedback condition, writers revised their first drafts, and the same reviewers reviewed the final drafts. An expert reviewed all of the novices' writing, while individual peers reviewed writing from six peers.; Based on the expert evaluation, surprisingly, it was found that those receiving feedback from multiple peers showed a significant improvement over those receiving feedback from the expert. To trace the sources of the improvement, qualitative protocol analyses were performed on types of feedback from each source and types of revisions. In addition, an exploratory causal connection model was analyzed to explain the impact of feedback on the improvement of writing quality through revision. Alternative hypotheses and implications are discussed.