Social network analysis of student interaction in team-based learning

• Main Author: Lau, Vincent Wei Teck
• Other Authors: Erik Anders Mikael Gustavsson
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Science::Mathematics::Statistics; Science::General::Education
• Online Access: https://hdl.handle.net/10356/166439
• Institution: Nanyang Technological University

This study aimed to explore the role of student interactions in Team-Based Learning (TBL) by utilizing Social Network Analysis (SNA) to analyze student interactions in two undergraduate calculus modules at NTU. The research involved surveys after each tRAT exercise, with participants distributing 100 points among team members based on perceived contributions levels. Data cleaning, imputation, and network generation were performed to generate social network features, such as degree, betweenness, closeness and eigenvector centrality. A weak positive correlation was found between individual contribution ratings and individual performance, but with SNA features and weights transformation, a R2 value of 28.8% was achieved with only four predictors in a linear model. This study suggests that individual performance is positively correlated with pre-ability and participation in the collaborative nature of TBL, and teams with stronger collaboration performed strongly, possibly due to the co-creation of understanding within the team. This study also gives an example of how SNA was able to improve the accuracy of models and how weights transformation could be used to improve linear models involving SNA.