Exploring ChatGPT as a writing assessment tool

• Published in: Innovations in education and teaching international Vol. 62; no. 3; pp. 867 - 882
• Main Authors: Bucol, Junifer Leal, Sangkawong, Napattanissa
• Format: Journal Article
• Language: English
• Published: Routledge 04.05.2025
• Subjects: automated assessment; ChatGPT; corrective feedback; rubrics; writing
• ISSN: 1470-3297; 1470-3300
• DOI: 10.1080/14703297.2024.2363901

This research paper employs an exploratory framework to evaluate the potential of ChatGPT as an Automated Writing Evaluation (AWE) tool in teaching English as a Foreign Language (EFL) in Thailand. The main objective is to investigate how well ChatGPT can assess students' writing using prompts and pre-defined rubrics compared to human raters. Moreover, the study examines its strengths and weaknesses as an assessment tool by analysing the teachers' reflections during the assessment process. Quantitative analyses revealed significant relationships between trial accounts in comparison with the human ratings. Qualitative analysis unearths patterns in the feedback, shedding light on ChatGPT's strengths and its limitations as an AWE tool. ChatGPT displays substantial promise as an AWE tool, offering distinct features such as human-like interface, consistency, efficiency, and scalability. Nonetheless, educators must be cognisant of its limitations. This study recognises that the strategic use of ChatGPT could enhance the evaluation process among teachers and foster the development of EFL students' written communication skills.