Analyzing University Students’ Attitude and Behavior Toward AI Using the Extended Unified Theory of Acceptance and Use of Technology Model

Abstract

This quantitative study using Partial Least Square Structural Equation Modeling (PLS-SEM) examined a structural model of the attitudes and behaviors of university students toward AI in higher education. The results obtained using SmartPLS 4.0 indicate that the constructs exhibit validity and reliability (λ ≥ 0.708, α=0.767-0.948, AVE=0.584-0.777, HTMT=< 3.3). Further, the analysis of the hypothesized extended Unified Theory of Acceptance and Use (UTAUT) model reveals that AI Awareness significantly impacts Attitude toward AI (β = 0.156, p = 0.003) and Behavioral Intention to Use AI (BIU) (β = 0.337, p < 0.001). AI Trust also significantly influences Attitude toward AI (β = 0.366, p < 0.001) and BIU-AI (β = 0.173, p = 0.007). Additionally, Attitude toward AI is a strong predictor of BIU-AI (β = 0.457, p < 0.001). Social Influence significantly affects Attitude toward AI (β = 0.21, p < 0.001), while Effort Expectancy and Performance Expectancy do not show significant effects in this context. The link between Facilitating Conditions and BIU-AI is also insignificant. The model explained a substantial portion of the variance in attitude (R2 =0.612) and behavior (R2 =0.710). Fit indices indicate good model fit, and predictive relevance metrics were satisfactory.