Exploring the Predictors of AI Chatbot Usage Intensity Among Students: Within- and Between-Person Relationships Using the Technology Acceptance Model

Abstract

The current research investigated the factors associated with the intensity of AI chatbot usage among university students, applying the Technology Acceptance Model (TAM) and its extended version, TAM3. A daily diary study over five days was conducted among univserity students, distinguishing between inter-individual (between-person) and intra-individual (within-person) variations. Multilevel structural equation modeling (SEM) was used to analyze the data. In Study 1 (N = 72), results indicated that AI chatbot anxiety was associated with perceived ease of use (PEOU) and perceived usefulness (PU), which serially mediated the link with AI chatbot usage intensity. Study 2 (N = 153), supported these findings and further explored the roles of facilitating conditions and subjective norm as additional predictors of PEOU and PU. Results from both studies demonstrated that, at the between-person level, students with higher average levels of PEOU and PU reported more intensive AI chatbot usage. In Study 1, the relationship between PEOU and usage intensity was mediated through PU at the within-person level, while the mediation model was not supported in Study 2. Post-hoc comparisons highlighted much higher variability in PEOU and PU in Study 1 compared to Study 2. The results have practical implications for enhancing AI chatbot adoption in educational settings. Emphasizing user-friendly interfaces, reducing AI-related anxiety, providing robust technical support, and leveraging peer influence may enhance the usage intensity of AI chatbots. This study underscores the necessity of considering both stable individual differences and dynamic daily influences to better understand AI chatbot usage patterns among students.