LAUSR

Abstract Prior research on the complex process of revision based upon peer feedback has focused on characteristics of each piece of feedback in isolation. Multipeer feedback allows for feedback to be repeated (or not), which could be a signal of feedback quality or be especially persuasive to peers. Separately, little research has examined how well peers select more impactful and accurate peer feedback in their revisions, whether repeated or not. We analyzed almost 2,000 peer comments received by 107 students in a secondary writing course in the US to determine whether feedback quality and feedback frequency predicted feedback implementation. Controlling for other feedback features and context factors, students were much more likely to implement feedback as both feedback quality and feedback frequency increased, surprisingly with no interaction (i.e., even low-quality comments were more likely to be implemented when repeated). However, low-quality comments often partially overlapped with high-quality comments, providing a potential explanation for the lack of an interaction. Finally, consideration of feedback frequency and feedback quality provides new insights into which feedback features are actually related to implementation. The results generally allay concerns about the blind-leading-the-blind in peer feedback as well as pushing for peer feedback arrangements that produce more overlapping comments.