University students' engagement with generative AI-supported automated writing evaluation (AWE) feedback

Studies have shown that engagement with automated writing evaluation (AWE) feedback can have a positive impact on writing development. With the emergence of generative AI (GAI), there is potential to improve AWE feedback and its reception by writers, calling for research on how writers work with GAI-supported AWE feedback. This study therefore aims to examine students' behavioural, cognitive and affective engagement with such feedback through a GAI-powered AWE platform, Learnalytics. A qualitative, multiple-case study approach was adopted. Four undergraduate engineering students with diverse backgrounds and varying levels of writing proficiency at a university in Hong Kong were invited to use Learnalytics to complete and revise their drafts of a research report for capstone project courses. Data were collected through student drafts, computer screen recordings, stimulated recall sessions and semi-structured interviews, and were analysed qualitatively. Findings reveal how forms and levels of engagement were mediated by individual differences and various aspects of participants' writing process, experience and considerations. This study contributes to the growing body of research into the intersection of GAI and L2 writing. (174 words)