- Research Article
- 10.1080/17501229.2026.2654834
Beyond traditional modeling: a network analysis of university L2 learners’ attitudes, agency, acceptance, and engagement with AI-generated feedback
- Apr 09, 2026
- Innovation in Language Learning and Teaching
- Renqiang Wang + 2 more +2
ABSTRACT The rapid development of generative artificial intelligence (GenAI, shortened as AI thereafter) has reshaped feedback delivery and engagement in higher education contexts, yet it remains unknown how second or foreign (L2) learners’ attitudes toward AI, agency in using, and acceptance of AI connect with their engagement with AI-generated feedback. This study, therefore, utilized network analysis to investigate the connections among the domains of the above constructs among university students in order to find out central nodes and key bridging edges. Results indicated that learners’ ability and mentality, two dimensions of agency, served as two core psychological constructs linking cognitive, behavioral, and affective dimensions of engagement with AI-generated feedback, while attitudes towards AI were in a peripheral position within the network. Notably, agency emerged as a bridging node, with the strongest connection observed between action and cognitive engagement, followed by the connection between ability and behavioral engagement. Overall, students demonstrated a high level of engagement with AI-generated feedback and AI agency, which were intertwined with rather than independent of each other. Pedagogical implications are discussed.
Read more