In-class AI usage and university students' attitudes: The mediating role of perceived learning engagement benefits.

With the widespread adoption of artificial intelligence (AI), the usage of various AI tools by university students to support learning has become increasingly common in classroom settings. Focusing on routine in-class AI usage and learning engagement, this study explores the mediating role of perceived learning engagement benefits in linking students' AI usage behavior to their attitudes, and investigates whether its mediating effects differ across various dimensions of learning engagement. Based on data from 287 questionnaires, structural equation modeling and Bayesian analysis were used to test the proposed hypotheses. The results indicate that in-class AI usage was associated with students' attitudes both directly and also indirectly by enhancing cognitive engagement and emotional engagement, both of which exhibited comparable mediating effects. In general, the university students exhibited positive attitudes toward in-class AI usage, and these attitudes were positively associated with their perceived learning engagement benefits. These results suggest that the value of AI in classroom settings may relate not only to learning efficiency, but also to students' learning experiences, particularly in terms of perceived cognitive and emotional engagement benefits. Notably, although in-class AI usage was positively associated with all dimensions of perceived engagement benefits, perceived behavioral engagement benefits showed no significant indirect effect on attitudes. This null finding warrants further exploration in future research.