Factors Influencing Vocational Students' Acceptance of Generative AI for Learning: A PLS-SEM Approach

Abstract

The present study aimed to: (1) examine the levels of perception and opinion of vocational students toward factors influencing their acceptance and use of Generative AI (GenAI) for learning; (2) develop and validate a structural model of causal relationships among such factors; and (3) analyze the direct, indirect, and total effects of predictor variables on behavioral intention and actual use behavior. A sample of 500 students enrolled in Vocational Certificate (VC) and Higher Vocational Certificate (HVC) programs under the Office of the Vocational Education Commission was selected through stratified random sampling. Data was collected using a seven-point Likert scale questionnaire (reliability = 0.978) and analyzed via descriptive statistics and Partial Least Squares Structural Equation Modeling (PLS-SEM) using ADANCO software. Results revealed that vocational students held highly positive opinions regarding GenAI acceptance across all dimensions. The PLS-SEM analysis confirmed satisfactory model fit. Performance expectancy (β = 0.149, p < 0.05), effort expectancy (β = 0.138, p < 0.05), social influence (β = 0.127, p < 0.05), hedonic motivation (β = 0.194, p < 0.01), personal innovativeness (β = 0.262, p < 0.01), and trust (β = 0.147, p < 0.01) all had significant positive effects on behavioral intention. Privacy did not exert a significant effect on behavioral intention. Behavioral intention strongly predicted actual use behavior (β = 0.625, p < 0.01), with the model explaining 79.97% of variance in behavioral intention and 39.05% in use behavior. Findings highlight the growing importance of GenAI in vocational education and suggest that institutions should promote AI literacy, ethical AI practices, and supportive learning environments.