Artificial Intelligence Assisted Feedback in EFL Writing: Student Perceptions of ChatGPT in an Indonesian Islamic University

Abstract

The growing presence of artificial intelligence in education has introduced new possibilities for providing feedback in academic writing. This study investigates students’ perceptions of feedback generated by ChatGPT in an argumentative writing course in the English Language Education Program at Universitas Islam Negeri Mataram, Indonesia. The research aims to understand how students perceive the usefulness of ChatGPT feedback, identify the aspects of feedback they find most helpful, and examine the challenges they experience when using the tool during the writing revision process. The study employed a descriptive mixed-method approach involving thirty-five undergraduate students. Data were collected through a rubric-based writing assessment, a perception questionnaire, and semi-structured interviews. The quantitative data were analyzed using descriptive statistics, while qualitative responses were examined through thematic analysis. The findings indicate that students generally perceive ChatGPT feedback as useful in improving grammar accuracy, sentence clarity, and organization of ideas in their essays. At the same time, students reported that some suggestions required careful evaluation because they did not always align with the intended argument. The study suggests that ChatGPT can function as a supplementary feedback resource that supports students during the revision stage. When combined with instructor guidance, artificial intelligence tools may contribute to more interactive and reflective learning environments in academic writing courses within Islamic higher education.