EFFECTS OF AI-ASSISTED FEEDBACK ON THE WRITING ACCURACY AND LEARNER AUTONOMY OF FIRST-YEAR EFL STUDENTS: A QUASI-EXPERIMENTAL STUDY IN AN UZBEK UNIVERSITY CONTEXT

Generative artificial intelligence (AI) tools are increasingly available to language learners, yet controlled evidence on their pedagogical value in Central Asian university classrooms remains scarce. This quasi-experimental study examined the effects of AI-assisted feedback on the writing accuracy and learner autonomy of first-year students of English as a foreign language (EFL) at a state university in Uzbekistan. Sixty students aged 18–20, taught in two intact classes, took part in an eight-week intervention. The experimental group (n = 30) received AI-assisted feedback supported by teacher guidance, whereas the comparison group (n = 30) received conventional teacher feedback. Writing accuracy was measured with pre- and post-tests scored on an analytic rubric, and learner autonomy with a 12-item questionnaire. Data were analysed with paired and independent-samples t-tests and Cohen’s d. The experimental group outperformed the comparison group in post-test accuracy and in gains on both measures, with large effect sizes. The findings are discussed in light of feedback, noticing, and self-regulation theory, and implications for the responsible integration of AI feedback into EFL writing instruction are outlined.