Comparing the Effects of ChatGPT-Assisted vs. Teacher Feedback Writing Practice on the Accuracy and Complexity of Intermediate Iranian EFL Learners' Writing

سال انتشار: 1405
نوع سند: مقاله کنفرانسی
زبان: انگلیسی
مشاهده: 23

فایل این مقاله در 15 صفحه با فرمت PDF قابل دریافت می باشد

استخراج به نرم افزارهای پژوهشی:

لینک ثابت به این مقاله:

شناسه ملی سند علمی:

AELEI01_030

تاریخ نمایه سازی: 22 شهریور 1405

چکیده مقاله:

This study examined the comparative effects of ChatGPT-assisted feedback and teacher feedback on the grammatical accuracy and syntactic complexity of intermediate Iranian EFL learners’ argumentative writing. Seventy-five learners, homogenized through the Oxford Placement Test, were randomly assigned to a ChatGPT-assisted feedback group, a teacher-feedback group, and a control group. Using a quasi-experimental pretest–posttest design, participants completed two argumentative writing tasks scored with an adapted IELTS Writing Task ۲ rubric emphasizing accuracy and complexity. Over an eight-session intervention, the experimental groups received either AI-generated feedback via standardized prompts or teacher-provided corrective feedback, while the control group completed the same writing tasks without systematic feedback. Nonparametric analyses indicated that both experimental groups significantly outperformed the control group in accuracy and complexity, demonstrating the effectiveness of both feedback types. However, no statistically significant difference emerged between the two experimental groups, suggesting that ChatGPT-assisted feedback was as effective as teacher feedback in improving learners’ written performance. These findings highlight the potential of AI-based feedback as a viable supplement to traditional teacher feedback, particularly in contexts with limited instructional time or large class sizes. Pedagogical implications include integrating ChatGPT to support revision practices, enhance learner autonomy, and complement teacher feedback. Limitations include the short duration of the intervention, the use of a single rater, and the focus on only two dimensions of writing performance. Future research should explore long-term effects, multiple writing dimensions, and varied proficiency levels.

نویسندگان

Mohammad Amir Sinafar

Phd Candidate in TEFL, Department of English Language, CT.C., Islamic Azad University, Tehran, Iran

Mehrdad Rezaee۲

Assistant Professor of Applied Linguistics, Department of English Language, CT.C., Islamic Azad University, Tehran, Iran