The Impact of AI Conversational Agents on Learning Performance and Experience in Desktop VR Environments: An Empirical Study

• In Event: The Role of Artificial Intelligence in Transforming Learning, Assessment, and Educational Practices

Abstract

AI Conversational Agents (AICA) can create generative learning experiences by addressing the rigidity of fixed program settings in VR-based learning. This study employed a quasi-experimental design to investigate AICA's impact on performance, motivation, engagement, flow, and cognitive load in a desktop VR setting. Sixty students were divided into an AICA-guided group (n=30) and a control group (n=30) without guidance. Results showed that the AICA group outperformed the control group in academic performance, motivation, engagement, and flow, with no significant difference in cognitive load. The frequency of questions asked to AICA positively correlated with post-test scores but negatively with motivation and flow. This study extends generative learning theory and provides empirical support for AICA's effectiveness in DVR learning environments.

Authors

• Xinyi Wang, East China Normal University

• Shaojian Xi, East China Normal University

• Chenpeng Yuan, East China Normal University

• Xiaoqing Gu, East China Normal University