Student learning agency, encompassing key abilities, essential mental characteristics and active actions, is recognized as a crucial factor for effective teaching and learning in the generative artificial intelligence (AI) era. This study examines the structural relationships among perceptions of generative AI (i.e., effort expectancy, performance expectancy, social influence, and facilitating conditions), behavioral intention, use experience, and student learning agency by integrating sociocultural perceptions, intentionality, and life course perspectives. Data were collected from 765 Chinese university students and analyzed using confirmatory factor analysis and structural equation modeling. Results showed that key abilities and essential mental characteristics predicted active actions, while key abilities predicted essential mental characteristics; behavioral intention predicted active actions and is predicted by key abilities; performance expectancy and social influence predicted behavioral intention; performance expectancy predicted essential mental characteristics; facilitating conditions predicted key abilities but negatively predicted active actions; the frequency of generative AI use predicted key abilities but negatively predicted essential mental characteristics; the number of scenarios of generative AI use predicted active actions and key abilities; use experience with generative AI negatively moderated the social influence–behavioral intention relationship. These findings provided valuable implications for designing educational strategies to enhance student learning agency in generative AI-supported environments.
Xia et al. (2025) studied this question.