Claim Missing Document
Check
Articles

Found 1 Documents
Search
Journal : JOLLT Journal of Languages and Language Teaching

ChatGPT-Based Reflective Feedback to Improve EFL Graduate Students’ Academic Writing for Publication Nurjati, Nunung; Syahria, Nukmatus; Khan, Aslam Khan Bin Samahs
Journal of Languages and Language Teaching Vol. 13 No. 4 (2025): October
Publisher : Universitas Pendidikan Mandalika

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33394/jollt.v13i4.14858

Abstract

This study investigates the use of ChatGPT-based reflective feedback to enhance academic writing among graduate's students in English as a Foreign Language (EFL) programs. With the increasing interest in AI tools for improving writing skills, the research examines ChatGPT's effectiveness in providing personalized, immediate feedback on students' drafts. Utilizing a case study design, the study focuses on 20 graduate graduate students and employs pre- and post-writing tests, reflective journals, qualitative surveys, and interviews to assess the impact of ChatGPT feedback on writing performance and perceptions. The findings indicate significant improvements in writing quality, particularly in areas such as essay organization, grammar, vocabulary, argumentation, coherence, clarity, and engagement after integrating ChatGPT feedback. Descriptive comparisons revealed noticeable gains in average scores, such as grammar (from 78 to 88), coherence and flow (66 to 87), and clarity and conciseness (60 to 80), with more modest gains in argumentation and critical analysis. While reflective journals and interviews revealed that students found the feedback process empowering, fostering greater learner autonomy and self-regulation. Students expressed appreciation for the immediate nature of the feedback, which allowed them to make timely revisions. However, challenges arose in interpreting and applying complex feedback, especially concerning higher-order writing skills like argumentation and critical analysis. The study suggests pedagogical implications for integrating AI feedback in EFL classrooms, emphasizing a hybrid feedback model that combines AI and human input to meet diverse learner needs. Further research with individual-level data and statistical testing is recommended to validate these outcomes. These findings highlight the potential of AI-assisted tools like ChatGPT to complement human instruction, personalize learning, and support the development of academic writing skills in diverse educational settings.