Abstract
Despite the increasing interest in AI technologies in education, there is a gap in understanding the factors influencing the adoption of ChatGPT among Nigerian higher education students. Research has not comprehensively explored these factors in the Nigerian context, leaving a significant gap in understanding technology adoption in this setting. This study addressed this gap by investigating the predictors of students' behavioural intentions and actual usage behaviour regarding ChatGPT using the UTAUT2 framework. A cross-sectional correlational research design was used in this study. A sample of 8,496 higher education students from diverse institutions in Nigeria participated in the study. The data were collected using the Higher Education Students’ ChatGPT Utilisation Questionnaire (HESCUQ). The findings reveal several significant predictors of students' behavioural intentions and actual usage of the ChatGPT. Performance expectancy, social influence, hedonic motivation, and habit emerged as significant positive predictors of behavioural intention, while effort expectancy and facilitating conditions were significant negative predictors. For ChatGPT use behaviour, facilitating conditions, hedonic motivation, habit, and behavioural intention were significant positive predictors, whereas performance expectancy and social influence were significant negative predictors. Behavioural intention mediated the relationships between several factors and ChatGPT usage behaviour: positively for some (PE, SI, HM, and HB) and negatively for others (EE and FC). This study contributes to understanding the adoption of the ChatGPT in higher education contexts.
Keywords
Artificial intelligence, chatbots, large language models, technology use, structural equation modelling.
