AI 피드백에 대한 학습자의 수용 판단과 다시 쓰기 반영 양상 연구 - 중급 단계의 학문 목적 한국어 학습자를 대상으로

초록

This study examines how intermediate learners of Korean for academic purposes judge and implement AI feedback in rewriting. The participants were 30 intermediate learners enrolled in an academic Korean writing course at a university in Seoul. The data comprised first drafts, revised drafts, and feedback uptake judgment forms. The analysis examined changes between drafts, learners’ uptake or non-uptake judgments and reasons, and the actual implementation of each feedback item as uptake, partial uptake, non-uptake, or misuptake. The findings showed that the learners’ revised drafts tended to improve in terms of content, organization, and expression. Learners judged 152 out of 167 AI feedback items as acceptable, indicating a high self-reported uptake rate. However, the actual implementation rate was lower than the learners’ self-reported uptake rate. In particular, non-uptake and misuptake were observed mainly in content and organization. These cases appeared to result from learners’ insufficient understanding of the focus and intention of the feedback and their difficulty in applying it to the argumentative structure of their own writing. The findings suggest that AI feedback uptake should be examined not only as learners’ willingness to accept feedback but also as their ability to interpret, judge, and implement it in revision. This study highlights the need to develop learners’ feedback literacy in AI-supported Korean writing instruction for academic purposes.

키워드