Balancing support and autonomy

Abstract

The integration of Artificial Intelligence (AI) into English as a Foreign Language (EFL) writing provides students with new ways to improve the quality of their academic writing. This study investigated how students interacted with ChatGPT feedback when revising academic essays. The essays used in this study were originally written by the participants for their Writing IV (Advanced Writing) course in the fourth semester. The essays represented different academic types, including argumentative, cause-effect, and persuasive essays. A qualitative case study design was employed involving three English Education students from different semesters who had previously completed Writing I, II, III, and IV courses, and selected through purposive sampling. The students accessed ChatGPT through their personal accounts using GPT-4o to obtain feedback. Data were collected through semi-structured interviews and documentation of essay drafts (before & after revision). The results showed that ChatGPT provided broad suggestions, specifically helping students improve grammar, vocabulary, content development, arguments, and sentence organization. However, some feedback was considered too formal, complex, or inappropriate for the students' intentions, so critical assessment was needed in selecting which suggestions to use. The effectiveness of using ChatGPT depended on the students' ability to interact critically with the AI-generated feedback.

Ähnliche Arbeiten