<![CDATA[This study aims to analyze the influence of personalized AI feedback on writing accuracy and learner autonomy of EFL students in universities. The study used a quantitative approach with a quasi-experimental pretest-posttest control group design involving 120 students, consisting of 60 experimental group students and 60 control groups. The experimental group obtained AI-based feedback, while the control group obtained conventional feedback. Data were analyzed using descriptive statistics, ANCOVA, regression, and mediation analysis. The results showed that personalized AI feedback had a significant effect on writing accuracy (F(1.117) = 74.98, p < .001, partial η² = .390) and learner autonomy (F(1.117) = 72.38, p < .001, partial η² = .382). In addition, learner autonomy has a positive effect on writing accuracy (β = .61, p < .001) and mediates the effect of personalized AI feedback on writing accuracy (β = .32; 95% CI [0.189, 0.469]). These findings show that AI feedback not only helps improve writing accuracy, but also encourages student independence in monitoring, evaluating, and revising writing. Therefore, the application of AI in writing learning needs to be designed to support student engagement and independent decision-making.]]>
Copyrights © 2026
