An investigation into the use of Rasch analysis to aid L2 writers in anonymized peer-assisted learning

Article appearing in Shiken 24.2 (December 2020) pp. 39-53; https://doi.org/10.37546/JALTSIG.TEVAL24.2-3

Jeffrey Martin
J. F. Oberlin University

Abstract

This study critically evaluated an anonymized peer feedback and assessment design for L2 writers enhanced by the use of Rasch analysis. This approach centered on the acts of giving assessment (Topping, 1998) and feedback (Lundstrom & Baker, 2009) in exchange with multiple peers. Each participant received feedback comments and class-wide statistical measures summarized for students without the need for their peers to rate all papers. Anonymity was maintained to bring unencumbered attention to the role of the reader (Booth et al., 2008) and to provide a space for interpretation and reflection on the potentially contrasting data and experiences that emerge. This process is argued to drive cognitive development and improve L2 writing skills. An initial trial with 15 high-proficiency EFL learners indicated that the design facilitated an effective exchange for each participant. The effects of anonymously including teacher comments also brought informative insights about the perception of feedback and its sources. Issues were found regarding overly narrow use of the ratings scales by some participants. A 6-point rating scale is proposed for more differentiated scoring. Overall, positive engagement and reception by the participants suggests that this peer assisted learning approach holds promise for L2 writers.

Keywords: peer feedback, peer assessment, L2 writing, anonymous feedback/assessment, Rasch analysis, judging plan

Download full article (PDF)