UNLOCKING STUDENT ENGAGEMENT WITH GENERATIVE AI: NEED SATISFACTION AS THE KEY, SELF-EFFICACY AS THE GATEKEEPER

Abstract

Understanding how learning environments support student engagement requires attention to the motivational processes through which learners internalize instructional resources. Drawing on Self-Determination Theory, this study examines whether perceived affordances of AI-supported learning environments promote students’ classroom engagement through psychological need satisfaction, and whether students’ AI self-efficacy conditions this process. Survey data were collected from 389 undergraduates and analyzed using structural equation modeling. The results indicate that perceived AI affordances positively predict behavioral, cognitive, and emotional engagement indirectly via enhanced need satisfaction, with the strongest indirect effect observed for cognitive engagement (β = 0.385). In addition, AI self-efficacy modestly strengthened the relationship between perceived affordances and need satisfaction, suggesting a meaningful boundary condition in the motivational process. These findings clarify how and for whom supportive learning environments foster engagement, highlighting the central role of psychological need satisfaction in translating contextual resources into sustained classroom involvement. From an instructional perspective, the study highlights the importance of designing learning environments, rather than focusing solely on technologies, that effectively support students’ autonomy, competence, and relatedness.