Investigating the Factors Affecting University Students' Adoption of Generative AI through PLS-SEM and Machine Learning

Abstract

Research on behavioral intentions to adopt generative artificial intelligence (AI) has only emerged in recent years and has mainly focused on Western, African, Middle Eastern, or other non-Chinese regions, leaving the unique context of Chinese higher education relatively underexplored. This study aims to examine the facilitating, inhibiting, and moderating factors influencing Chinese university students' use of generative AI for learning purposes. The empirical results of the PLS-SEM model indicate that facilitating factors such as performance expectancy, hedonic motivation, habit, perceived trust, and personal innovativeness significantly affect Chinese university students' behavioral intention to use generative AI. Furthermore, the results of the machine learning CatBoost model and SHAP analysis reveal that the relative importance of predictors of Chinese university students' behavioral intention to use generative <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{A I}$</tex> is ranked as follows: habit, personal innovativeness, perceived trust, hedonic motivation, and performance expectancy. Due to the unique context of Chinese university students, inhibiting factors such as academic guilt and privacy concerns do not exert significant influence on their behavioral intention to use generative AI. Gender moderates only the effect of price value, while no moderating effects are observed on other facilitating or inhibiting factors. This study expands the scope of technology adoption research.

Ähnliche Arbeiten