Interpreting peer assessment: an algorithm to navigate group problem detection

Abstract

When graduates enter the workforce, they are often expected to be able to successfully collaborate in a team. For this reason, universities regularly incorporate group work in their teachings. However, in group work problems such as social loafing and poor common can occur. Also, teachers do not have enough time to keep an eye on all groups participating in their course project(s). This research explores how peer and self-assessment can be used to understand group dynamics and performance in higher education. Given the challenges of monitoring group work and identifying issues, the study aims to develop a method for detecting these problems early through computer-supported assessment tools. Three primary research methods were used: a literature review to identify key variables influencing group dynamics. Next, focus groups were held to gather insights from students and teachers on interpreting peer assessments. Lastly, a field study was organised to evaluate the effectiveness of the rule-based algorithm created for this study in real classroom settings. At the end of the field study, evaluation interviews were done with the teachers from the courses to evaluate the effectiveness, accuracy and usefulness of the tools. The findings highlight important individual- and group-related variables, such as leadership, prior teamwork experiences and grade ambitions, that shape how students assess each other. The study concludes that these variables provide important context when using structured assessment tools and the rule-based algorithm to understand group dynamics and performance. This approach can help teachers more effectively identify and address group challenges.