2025 ASEE Annual Conference & Exposition

This study looks into the use of team evaluation software, incorporating peer ratings, peer comments, and machine-learning-based analysis, to assess the project performance of student project teams. Teamwork is an essential competency for students. The early development of collaborative skills is critical for academic success and future career success.

Previous studies have suggested that the data-driven team evaluation could help with team performance evaluation. However, most of the team-based software will provides peer rating without detailed feedback of student team performance. CATME (Comprehensive Assessment of Team Member Effectiveness) greatly facilitates peer assessments by allowing students to rate and comment on each other's contributions, fostering accountability and constructive feedback. Additionally, machine learning algorithms analyze the collected data to identify patterns in team dynamics, on the five dimensions of CATME (Comprehensive Assessment of Team Member Effectiveness), individual participation and team synergy.

Natural Language Processing (NLP) tools plays a crucial role in evaluating team performance via analyzing communication, feedback, and interactions among team members. This study will further explore a variety of natural language processing tools, such as sentimental analysis, text classification, topic modeling, named entity recognition (identify potential project leaders), and keyword extraction.

By considering both human and machine evaluations, the study aims to provide a comprehensive assessment of team effectiveness, highlighting areas for growth for individual students. The findings suggest that this approach not only increase the efficiency of the evaluation process, but also possibly improve student engagement, and the overall quality of teamwork amongst student groups.