2025/07/26 by Torki, Roghaye, Rahimi Esfahani, Fariba, Kiyoumarsi, Farshad
Computer Science · Medicine · Social Sciences · #AI in Service Interactions #Artificial Intelligence #Artificial Intelligence in Healthcare and Education #Constructive Feedback #EFL Students #English Writing Skills #Optimization #Self-Regulated Learning #Writing and Handwriting Education
paper · doi:10.82553/josc.2025.202507261213086
openalex publication_date 2025/07/26 · openalex created_date 2025/12/19 · openalex updated_date 2026/07/01
Artificial Intelligence (AI) has emerged as a pivotal force in enhancing second language writing instruction by offering immediate, data-informed, and personalized feedback. This study investigates the effectiveness of AI-generated constructive feedback-via ChatGPT-in improving English as a Foreign Language (EFL) learners’ academic writing performance. Employing a quasi-experimental, mixed-methods design, 60 intermediate-level EFL students were divided into experimental and control groups. Over eight weeks, the experimental group received iterative feedback through ChatGPT, while the control group relied on traditional teacher feedback. Writing performance was assessed using an adapted IELTS Task 2 rubric. Quantitative results revealed substantial improvements in the experimental group’s grammar (+28%), coherence (+22%), lexical resource (+22%), and task achievement (+21%), with a large overall effect size (Cohen’s d = 0.72). Qualitative data from post-study surveys supported these findings, highlighting enhanced learner confidence and autonomy. The study proposes a Hybrid AI-Teacher Feedback Model that integrates AI precision with human pedagogical insight. This approach addresses both instructional efficiency and cultural appropriateness, offering a scalable and ethically grounded framework for modern writing pedagogy