Abstract:

This research work examines group dynamics of university students in cohorts. A cohort refers to a group of students that start, move through and complete a program together. Prior studies have found that this unique support system plays a critical role in students' learning. Hence, it is important for educators to have a good understanding of student group dynamics. This study presents a longitudinal study of student project team formation behavior over the course of five calendar years (2009 through 2014). We perform a social network analysis to identify fundamental network structures amongst student project teams and to identify characteristics of key individuals or influencers. We also draw correlation between team formation behavior and students' performance (e.g. course grade and GPA). Lastly, we examine the relationship between students' demographics (e.g. gender, age) and team formation behavior. We show that the network diameter decreases over time as students mix and mingle in forming project teams over the course of four years of their studies. We also show that due to the constraint of allowed group size (typically 4-5 per group) and due to the number of courses students take each semester, the diameter does not reduce drastically. Also, our analysis results show that the network density slowly decreases over time. As "cliques" form, students generally team up with familiar faces. Findings from this study are useful to the educators in understanding students' team formation behavior and how it relates to their performance.

Keywords:

Social network analysis, team formation, group dynamics.