Learner Use of AI Feedback in English Writing: Effects of AI Feedback in Writing according to Learner Traits

This study investigates how learners engage with AI feedback in English writing tasks and examines how learners’ initial writing proficiency, self-efficacy, and trust in AI feedback influence both writing performance and feedback usage behavior. The research involved approximately 40 university students with prior experience using AI-based feedback tools. Participants completed a diagnostic writing test, two questionnaires measuring writing self-efficacy and trust in AI feedback, four essay-writing tasks, and a post-writing test. Quantitative data were analyzed using SPSS 28.0 to examine relationships among learner variables and writing outcomes. The findings indicate that AI feedback alone does not guarantee improvement; rather, learners’ perceptions of feedback, motivational factors, and contextual guidance provided by instructors play a critical role in effective feedback uptake. In addition, the results suggest that structured instructional support can enhance learners’ ability to interpret and apply AI-generated feedback meaningfully. Based on these findings, the study offers practical implications for AI feedback system design, including the incorporation of emotionally sensitive language, contextual personalization, and educator-facilitated feedback interpretation. Future research should further investigate longitudinal effects and a wider range of text genres and learner populations to strengthen the pedagogical validity of AI-assisted feedback.