Towards AI-Assisted Immersive Learning:Factor Analysis of Learning Effect in K-CubeEdu-Metaverse
Authors: Ye Jia, Chen Li, Zackary P. T. Sin, Wang, Xiangzhi Eric, Jiongning Lian, Peter H. F. Ng, Xiao Huang, George Baciu, Cao, Jiannong, Qing Li
Published in: IEEE International Conference on Metaverse 2025 (2025)
Abstract
This study examines the impact of an AI-powered Virtual Teaching Assistant (NivTA) within a VR-based Edu-Metaverse (K-Cube), highlighting the roles of social presence, trust, and engagement in shaping learning outcomes. Grounded in Social Presence Theory, the Uses and Gratifications framework, and the Cognitive-Affective Theory of Learning with Media (CASTLE), our AI-Assisted Immersive Learning Framework emphasizes both cognitive and affective dimensions. A user study with 21 participants in a Cave Automatic Virtual Environment (CAVE) setting collected quantitative and qualitative data on trust, social presence, engagement, workload, and learning performance. Partial Least Squares Structural Equation Modeling revealed that heightened social presence fosters trust, which in turn drives behavioral, cognitive, and affective engagement. Notably, cognitive social presence was directly linked to better knowledge test scores, while confidence in test responses stemmed primarily from all forms of engagement. Overall, these findings underscore the significance of nurturing trust and social presence to enhance learner engagement and outcomes in AI-driven immersive educational environments.
