Harnessing Generative AI to Enhance Feedback Quality in Peer Evaluations within Project-Based Learning Contexts

This paper explores using generative artificial intelligence (AI) large language models (LLMs) to augment traditional peer feedback processes within project-based learning (PBL) environments where students work in teams. Our study contributes to the expanding digital educational frontier by harnessing the power of LLMs for summarizing and enhancing self and peer feedback. We present a novel educational innovation that utilizes GPT-4 to effectively summarize self and peer comments for student teams in PBL courses. Our research involved the development of a specific prompt grounded in relevant educational theories and the analysis of 118 AI-summarized Performance Feedback Reports that were generated for students. The findings reveal that the LLM significantly improves the quality of peer feedback by making it more constructive and actionable, addresses off-topic or inappropriate comments, and provides a meaningful comparison between a student’s self-evaluation and peer comments. Our paper underscores the importance of peer feedback systems as valuable pedagogical tools in PBL classes and demonstrates how AI tools can enhance these systems. The outcomes of this research serve as a foundational guide for faculty, especially those teaching in PBL environments with student teams, who are considering using AI to enhance teaching and learning. We demonstrate the potential of integrating these innovative practices in PBL classrooms to support students’ development of teamwork skills, marking a significant step towards a new digital frontier in engineering education. © 2024 TEMPUS Publications.