Abstract

Recent research in language education has focused on evaluating the efficacy of ChatGPT in language acquisition. This study explored the accuracy, clarity, and effectiveness of ChatGPT feedback and its effectiveness as an Automated Essay Scoring tool. Experienced EFL instructors were selected to identify the strengths and limitations of ChatGPT by comparing it to established criteria for writing feedback. The findings indicated that ChatGPT can support educators in delivering clear, constructive, and efficient feedback to students, particularly in the domains of grammar, vocabulary, and mechanics, as well as highlighting positive and negative aspects of students’ writings. However, instructor intervention remains necessary for addressing organizational and content-related errors. Moreover, unlike human instructors, ChatGPT appears to lack sensitivity to contextual factors such as learner interaction and proficiency level. This research endeavors to inform teachers, educators, artificial intelligence developers, and policymakers about the utility of ChatGPT; including its advantages and limitations, as a writing correction tool.

This work is licensed under a Creative Commons Attribution 4.0 License.