#1318: Analyzing Embodied Engagement in Co-located Collaboration through Kinesthetic Modality

This demo aims to explore the potential of pattern mining in the kinesthetic modality. Multimodal collaboration analytics (MMCA) is an emerging and expanding field in CSCL and LA, emphasizing its capability to capture the digital trace of embodied engagement. However, few studies have explored the kinesthetic modality, leaving theoretical and methodological frontiers of how to handle accelerometer data. To narrow the gap, we developed a multimodal sensor badge and proposed an analysis pipeline by adapting pattern mining to accelerometer data to examine embodied engagement in collaboration. Eight groups of 3 undergraduates participated in collaborative problem solving, and one high-performing and one low-performing group were analyzed as a case study. The results suggest that pattern mining reveals nuanced but meaningful differences in engagement that could not be identified using conventional methods. Our analysis pipeline suggests the possible direction for deriving meaningful insights from kinesthetic modality data and extending theories of collaboration.