A Novel Approach to Examine the Impact of Web-based Peer Review on the Revisions of L2 Writers

- The types of feedback, alteration, and the specific trait of feedback, recurring, are important predictors for revision.
- Alteration is the most 'simple' revision to implement as a clear suggestion is provided on how to change the text.
- Alteration might not always lead to an improvement of the text, if students do not consider them carefully, but blindly implement them.
- Web-based peer review systems might consider all peer reviews during the peer feedback process.
- The chosen method, Machine Learning, is effective when predicting implementation and invites replication studies.

The use of peer feedback to support student writing is increasingly popular among second language writing teachers. However, the effect peer feedback has on the revision and the writing process remains unclear. Technological advances have made the application of peer feedback in the L2 writing context more accessible. As a result, there is a growing body of research investigating web-based peer review and L2 writing. This study aims to better understand how L2 writers conduct peer feedback activities, by looking at the types and traits of the feedback and how they influence revisions made in subsequent drafts using a web-based peer review system. A new methodology for studying web-based peer review comments is introduced. The results suggest that a specific type of feedback, alteration, and specific type of feedback, recurring, are important predictors for revision.