|Abstract:||A prevailing question across studies has been whether and to what extent ESL writers can help each other when working in dyads or groups. While most research focused on the effectiveness of peer feedback, relatively few studies took interest in what L2 students actually do during a peer review. Moreover, there has been a growing awareness of and interest in paired and group oral assessment, especially in examining interactional competence during paired speaking tests; however, to date not much has been explored with regards to the nature of peer interaction during a writing test. This study proposes to fill this gap in SLA and L2 testing research by examining the peer-review section of the English Placement Test (EPT), a process-oriented, integrated writing placement test offered to newly-admitted international students at the University of Illinois. The current study examined (a) test-takers’ perceptions and attitudes towards the peer-review section in a testing context, (b) the nature of peer interaction during the peer feedback process, and (c) its potential link to test-takers’ writing performance.
Test-takers generally showed positive attitudes and perceptions towards the peer review. They appreciated the opportunity to receive feedback from their peers and perceived the activity as helpful in writing their final essay. Test-takers also indicated that they focused more on rhetorical features than lexico-grammar features during the peer review.
Four patterns of interaction were identified during the peer review: collaborative, dominant/dominant, dominant/dominant monologue, and passive/passive. The collaborative pattern was the most common among the EPT examinees (n=19, 60%) in line with the findings of the previous research on patterns of interaction in classroom settings (e.g., Storch 2002; Zheng, 2012). A new pattern not observed in prior classroom studies, which I named dominant/dominant monologue, was observed among 29% of the pairs (n=9), whose interaction was monologic in nature with little or no engagement.
Written texts produced by the test-takers were then triangulated with the post-test questionnaire responses and peer interaction data to observe what changes were made as a result of peer interaction. Collaborative pairs made more detailed, specific comments that were revision-oriented, and many of these suggestions were reflected in the final drafts. In contrast, comments exchanged between non-collaborative pairs were more global in nature, and therefore, it was difficult to identify specific edits reflected in the final drafts. Regardless, most non-collaborative pairs appreciated the opportunity to receive feedback from their peers, and many reported making changes based on their peers’ feedback on their post-test questionnaire.
This study’s results stand to have valuable implications for both second language writing and language testing at large, and they provide insight to the nature and effect of peer interaction in an authentic, academic writing placement test setting. This study also emphasizes the role of peer review training and how this can be used to facilitate peer interaction and help test-takers achieve their best possible writing performance.