Chinese Undergraduate L2 Writers' Metacognitive Regulation When Using Teacher and AI Feedback: A Qualitative Study
TL;DR
The reported metacognitive regulation of five Chinese undergraduate non-English majors who had experience using both teacher feedback and generative AI tools in L2 writing was explored, highlighting the importance of feedback literacy and learner agency in human-AI feedback environments.
Abstract
Generative artificial intelligence is increasingly used alongside teacher feedback in second-language writing, yet limited attention has been paid to how learners regulate their writing when engaging with these two feedback sources. This qualitative study explored the reported metacognitive regulation of five Chinese undergraduate non-English majors who had experience using both teacher feedback and generative AI tools in L2 writing. Data were collected through semi-structured interviews and analysed using reflexive thematic analysis. The findings indicated considerable variation in participants' planning, monitoring, checking and revision practices. Some participants described systematic regulation involving outlining, paragraph-level monitoring and revisions extending beyond explicitly marked problems, whereas others mainly corrected errors directly identified by teachers or AI tools. Most participants regarded teacher feedback as more authoritative for content, organisation, genre and assessment expectations, while AI feedback was valued primarily for its immediacy and support with grammar, vocabulary and sentence-level expression. Participants who reported clearer goals and more active monitoring also perceived greater benefits from feedback, although the study does not establish objectively measured improvement in writing ability. The findings highlight the importance of feedback literacy and learner agency in human-AI feedback environments. Pedagogically, students may benefit from explicit guidance on comparing, evaluating and selectively applying feedback from different sources.