Decoding Student Rating Behaviors: A Comparative Study of Frequent Self and Peer Assessments in Team Projects

Team projects are commonplace in higher education, including in the domain of software engineering, as they benefit students' collaborative learning and employability skills. However, persistent challenges (e.g., free riders, dominators, and conflicts) frequently occur in teams and, as a result, undermine teamwork effectiveness. To help identify these problems, self and peer assessments have been widely adopted and practiced by instructors, and several studies have examined students' rating behavior from those assessments. However, existing research has largely been limited to the analysis of one-time—usually at the end of a project—peer and self assessments. In our work, we have been exploring a lightweight, high-frequency peer and self assessment tool that is used throughout a project. In this paper, we present our analysis of 29,844 self and peer ratings obtained from 234 students who engaged in three team projects in a single course. Based on data collected from the participants of this study, we identified several findings related to students' rating behaviors and how instructors can make effective use of students' ratings from frequent peer assessments.