AI-Augmented Peer Led Team Learning for STEM Education

Active learning methodologies in higher education benefit students by reinforcing learning and critical skills during the class. In active pedagogical models such as Peer-Led Team Learning (PLTL) students have stronger course outcomes. The instructor is not present in a PLTL workshop and may not receive sufficient feedback from peer leaders. Additionally, these classes have large enrollments. There is a lack of AI-enabled tools that monitor or provide feedback during cPLTL workshops. The current study addresses this gap by proposing an AI-based multimodal solution using recordings of cyber Peer-Led Team Learning (cPLTL) classes. The machine learning model analyzes audio and text features to predict the outcome of a workshop. The results using multimodal learning show potential for further development of the tool. Such improved modeling will help reduce the instructor’s workload facilitating the integration of AI in education. This novel multimodal approach aims to enhance the student’s learning experience by providing an automated feedback mechanism to the instructor.