Student Engagement in AI ‐Supported Learning: Unpacking Psychological Drivers via Neural Network Analysis

ABSTRACT As artificial intelligence (AI) becomes increasingly embedded in higher education, student engagement in AI‐supported learning has become an important indicator of learning quality and educational effectiveness. However, existing research has paid relatively limited attention to the psychological mechanisms through which students become actively involved in AI‐supported learning environments. Grounded in control‐value theory, this study examined the relationships among growth mindset, enjoyment, hope, pride, and engagement in AI‐supported learning among university students. A questionnaire survey was conducted with 563 students who had prior experience of AI‐supported learning. Structural equation modelling was used to test the hypothesized relationships and mediating effects, while artificial neural network analysis was further employed to examine nonlinear predictive importance. The results showed that growth mindset was positively associated with enjoyment, hope, pride, and engagement. Enjoyment, hope, and pride were also positively associated with engagement and significantly mediated the relationship between growth mindset and engagement. The ANN results further indicated that pride was the strongest predictor of engagement, followed by growth mindset, enjoyment, and hope. These findings suggest that students' engagement in AI‐supported learning is associated not only with their beliefs about ability development, but also by their positive emotional experiences. The study contributes to educational research by offering an integrated explanatory and predictive account of student engagement in AI‐enhanced higher education.