The intelligent MMLA approach can provide useful feedback for the three presences and learning outcomes in the metaverse environment, which can inform pedagogical interventions in secondary school science education. 

This study aims to examine social, cognitive, and teaching presences and student learning outcomes in collaborative science learning in the metaverse. The community of inquiry (CoI) framework is used to support online collaborative learning processes in the virtual world. The CoI framework consists of three elements: social, cognitive, and teaching presences, which are generally related to learning outcomes. However, there are some challenges in measuring these presences in online collaborative learning, such as lack of unified approaches, heterogeneous data, intrusive data collection methods, and limited research on multimodal learning analytics (MMLA) in the metaverse. 

To address these challenges, this study will design and implement an intelligent MMLA approach in secondary school education. The metaverse platform – “Learningverse” developed by the principal investigator’s team will be used as the collaborative learning environment. Learningverse provides an immersive environment where learners can use avatars that mirror themselves to participate in collaborative learning activities on a common computer with a webcam. 

The study will adopt design-based research with four phases: (1) analysis of needs and challenges, (2) design and development of an intelligent MMLA approach, (3) three iterative cycles of implementing and refining the approach to examine patterns of the three presences and learning outcomes, and (4) reflection. The participants are eight science teachers and eight classes of seventh graders. Data collection includes multimodal data logged in Learningverse, teacher and student retrospective interviews, and domain tests. Both qualitative and quantitative analysis methods are involved to identify specific indicators of each component of the three presences and their patterns, and the relationships between these patterns and learning outcomes. 

This study has both theoretical and practical significance. It is a pioneer study on developing an intelligent MMLA approach to identify specific indicators of social presence, cognitive and teaching presences, and their patterns in a 3D immersive and collaborative metaverse environment based on the CoI framework. The intelligent MMLA approach can provide useful feedback for the three presences and learning outcomes in the metaverse environment, which can inform pedagogical interventions in secondary school science education. Moreover, it has the potential to be scaled up to promote better immersive and collaborative science learning experiences in the metaverse at a larger scale both locally and internationally.