Assessing the quality of the collaborative learning process in iVR

Immersive virtual reality (iVR) has emerged as a promising tool in science education, with research documenting the potential benefits of collaborative learning in these environments. This study explores the dynamics of dyadic collaboration in iVR and presents a categorical framework to assess its quality. The framework was informed by previous works on collaborative learning process analysis and iVR collaborative learning theory. A qualitative multimodal design was used to document and compare patterns of social interaction. This design focused on describing students' verbal and non-verbal behaviours by analyzing video data, students’ interaction transcripts, and word count visualizations, allowing interactional patterns to be related to collaborative profiles. Nine undergraduate student pairs participated in an iVR learning experience to learn about water crystallization and the formation of snowflakes. Data included synchronized video recordings, word count, and transcripts of verbal exchanges. The framework differentiates collaboration quality into three groups: high, moderate, and low. These ratings are based on key aspects of collaboration such as verbal responsiveness, peer assistance, and turn-taking, but also incorporate unique iVR affordances like gestures, body movements, and object manipulation. The proposed framework provides a structured approach to understanding how students interact with others inside iVR and highlights the role of iVR in shaping collaborative dynamics. The findings contribute to theory by linking iVR affordances to collaborative processes and offer practical guidance for educators and designers.