Assessing Student Acceptance of an LLM-Integrated VR Public Speaking Simulation via Extended UTAUT

Abstract

Public speaking anxiety (PSA) is a common issue among the youth due to a lack of confidence, fear of strict evaluation, and concerns about self-image. This can cause negative effects, such as shyness, reduced social skills, and limited career goals. Past researchers asserted that virtual reality (VR) simulation with a virtual public speaking environment for practicing increases confidence. Moreover, other researchers applied artificial intelligence (AI) to enhance communication and foster student learning. To our understanding, there are currently few studies on integrating large language models (LLMs) in VR within the public speaking realm. In addition, analysis of behavioral intention to use the simulation is also essential before widely applying new technologies in educational environments. Therefore, we conducted a study based on the Unified Theory of Acceptance and Use of Technology (UTAUT) approach, using an LLM-integrated VR simulation system. In this study, we recruited 30 students from two public speaking sections and asked them to experience and rate the system. Results revealed that effort expectancy, facilitating conditions, and hedonic motivation are significant predictors of behavioral intention. Gender and prior VR experience were not moderating factors for connections between UTAUT constructs and behavioral intention. In addition, our results also highlighted the critical role of two antecedent variables: academic major and GPA levels, which had a significant effect on performance expectancy, effort expectancy, and hedonic motivation. Our research findings offer beneficial theoretical and practical implications for applying novel technologies to students.