Unpacking the Effects of Role Assignment on Interaction Processes in Socially Shared Regulation of Learning: A Multi-Method Approach

Abstract

Ineffective interaction processes often hinder the success of socially shared regulation of learning (SSRL) in online settings. While role assignment is a common scaffolding strategy in online collaborative learning, its influence on SSRL-based interactions remains underexplored. This study employed a 10-week quasi-experimental design to compare a role-assigned group with a non-role group (N = 40 undergraduates). Data sources included questionnaires analyzed via T-test and chat transcripts examined through qualitative coding, epistemic network analysis (ENA), and lag sequential analysis (LSA). Results indicated that role yielded significantly higher frequencies and complexity in cognitive (χ² = 151.334, p < .001), social (χ² = 97.05, p < .001), and teaching interactions (χ² = 171.72, p < .001). Although the overall volume of emotional interaction did not differ, the role-assigned group showed significantly more shifts from negative to positive emotions. Interaction patterns varied by role, with the task leader dominating teaching episodes and the emotion regulator triggering the most positive affective shifts. These findings suggest that aligning task scripts with specific SSRL roles can foster deeper shared regulation. We provide evidence-based prompts and monitoring checkpoints, supported by computational analyses, to assist teachers and instructional designers in enhancing online collaborative learning within educational computing environments.