摘要：Though corrective feedback (CF) has generally been accepted as an effective means for improving student writing, some debate still exists as to whether focused (narrow) or unfocused (broad) CF is more effective in improving student writing in English as a Foreign Language (EFL) contexts. A comparative study was undertaken of two groups of high-proficiency Japanese and international students of English at one private university in the Kansai area of Japan. A third control group who did not partake in any corrective feedback were also used for comparison. Both groups of students wrote argumentative essays on a subject of their choosing over the course of eight weeks. The first group, Treatment Group A, was comprised of seven Japanese and non-Japanese university students ( n = 7) who were trained in giving meta-linguistic (error coded), computer-mediated unfocused peer CF. The second group, Treatment Group B, was comprised of seven Japanese university students ( n = 7) who were trained in giving meta-linguistic, computer-mediated focused feedback on five errors identified as being the most common in an initial diagnostic writing sample done in the first week. The initial draft, post-peer CF draft, post-teacher CF draft, and final draft were then analyzed. Students’ ability to correctly resolve errors, and the number of errors per 100 words that emerged in each draft were then examined. The results suggest that unfocused peer and teacher CF may be a more effective means of reducing student errors in writing, possibly because it provides more overall learning opportunities.
