#1109: Using Multimodal Analytics-Supported Reflective Assessment to Promote Undergraduates’ Productive Social-emotional Interaction

This quasi-experimental study investigated the effects of multimodal analytics-supported reflective assessment (MASRA) on undergraduates' collaborative knowledge advancement and socio-emotional interactions. The experimental group (n = 48), utilizing MASRA with visualized analytics and reflective prompts, was compared to a comparison class (n = 38) using portfolio assessment. Analysis of 4034 discussion notes revealed the MASRA group significantly outperformed the comparison class across six collaboration dimensions, achieving nearly double the collective knowledge advancement. The experimental class demonstrated more positive and fewer negative socio-emotional interactions, alongside more frequent “negative → positive” transitions. Furthermore, MASRA strengthened the co-occurrence of metacognitive discourse and positive socio-emotional dynamics, suggesting it enhances students' ability to regulate emotions and maintain group cohesion. These findings indicate that MASRA is a scalable model that effectively fosters both cognitive-level knowledge innovation and productive socio-emotional dynamics in collaborative learning environments, offering evidence-based strategies for higher education.