Design instructors are integrating the use of online peer review platforms to keep pace with growing class sizes. However, these platforms typically prioritize randomized peer assignment strategies and show only the current solution to peers when writing feedback. This can result in low quality feedback in project-based design courses. We report on an experiment in which students (N=59) worked on twelve-week design projects and both wrote and received online feedback at four stages. The experiment tested a novel assignment strategy of peer mentorship, where peers were assigned to give feedback to all stages of the same project, and tested showing the context from the preceding design stage when composing feedback. The results showed that displaying the context from the preceding design stage led to feedback with higher perceived quality at the late design stages (but not at earlier stages) and feedback from mentors prompts longer responses from the feedback recipients. Our work contributes deeper empirical understanding of how assignment strategies and showing additional context affects peer feedback and provides practical guidelines for instructors to implement these methods in design courses.
|Number of pages||11|
|Journal||Proceedings of the ACM on Human-Computer Interaction|
|Issue number||CSCW, November 2017|
|Publication status||Published - 6 Dec 2017|