Generative Artificial Intelligence Use and Learning Engagement Among Medical Undergraduates: Statistical Indirect and Configurational Associations Involving Basic Psychological Need Satisfaction

As generative artificial intelligence (GAI) becomes increasingly integrated into undergraduate medical education, its relationship with students’ learning engagement warrants further investigation. Guided by self-determination theory, this cross-sectional questionnaire study surveyed 498 medical undergraduates at a medical university in China and used structural equation modeling (SEM) and fuzzy-set qualitative comparative analysis (fsQCA) to examine average associations and configurational patterns among GAI use, basic psychological need satisfaction, and learning engagement. SEM results showed that GAI use was positively but weakly associated with basic psychological need satisfaction, whereas basic psychological need satisfaction was more strongly associated with learning engagement. The direct association between GAI use and learning engagement was not statistically significant, while the statistical indirect association through basic psychological need satisfaction was significant but small. The fsQCA results showed that high GAI use was not a necessary condition for high learning engagement. The configuration combining autonomy, competence, and relatedness need satisfaction had the highest coverage and showed relatively stable results across threshold adjustments, with competence need satisfaction as a core condition. These findings suggest that the educational relevance of GAI use may be better understood in relation to students’ psychological need conditions than to use frequency alone.