Unveiling Dynamics of Student Engagement in Collaborative Discourse: A Temporal Clustering Approach (Poster 11)

• In Event: AERA Poster Session Thursday 1:45 p.m.
  In Poster Session: Advanced Technologies for Learning Poster Session 1

Abstract

This study examines the within-person dynamics and trajectories of student engagement in collaborative discourse through temporal clustering. Using an integrated analytical approach of natural language processing, network analysis, and clustering analysis, this study analyzed 810 annotations from 20 graduate students over seven weeks collected in a social annotation tool Perusall. The analysis revealed three distinct engagement profiles: less active but interaction-focused, balanced engagers, and end-of-semester workers. These profiles highlight the diverse engagement patterns and their evolution over time, emphasizing the need for pedagogical strategies that account for both individual differences (heterogeneity) and mutual understanding (intersubjectivity) in collaborative discourse learning. The findings offer actionable insights for designing effective collaborative learning environments, advancing the methodological approaches to analyzing discourse dynamics.

Authors

• Yeonji Jung, Texas A&M University

• Xinran Zhu, University of Illinois at Urbana-Champaign