Using epistemic network analysis to examine the cognitive engagement process in online collaborative learning

Cognitive engagement is essential for meaningful learning and is marked by interdependence and periodicity. However, the dynamic process and temporal aspects of cognitive engagement in collaborative learning have not been fully explored. This study investigates the cognitive engagement processes of different performance groups in online collaborative learning by analyzing discourse data from three activities. Content analysis and epistemic network analysis revealed that high-performing groups exhibited more task-regulated behaviors and stronger co-occurrences between perspective expression and perspective elaboration, as well as between perspective expression and low-level regulation. In contrast, low-performing groups showed more co-occurrences between perspective expression and information sharing. Additionally, the high- and low- performing groups demonstrated different patterns of cognitive engagement at various task stages. These findings offer practical guidance for designing effective collaborative learning activities to enhance cognitive engagement.