Speakers
Description
With the rapid advancement of Artificial Intelligence (AI), its integration into language assessment has gained growing popularity in recent years. This study examines the reliability of OpenAI’s language model, ChatGPT 4.0 in grading essays from the Writing Task 1 of The International English Language Testing System exam, assessing its consistency and alignment with official human graders. A quantitative approach is employed, including a comparison of mean scores, reliability measures such as intraclass correlation, and Bland-Altman analysis to compare grades from official examiners and ChatGPT across 60 essay answers. The results indicate that ChatGPT generally produces more conservative and consistent scores, while human raters tend to demonstrate greater variability and a broader scoring range. Although a moderate level of agreement is observed, the study identifies systematic differences in grading patterns, particularly in handling extreme performance levels. These discrepancies highlight the limitations of AI-generated scoring when used independently. Nevertheless, ChatGPT shows potential as a supplementary tool in language assessment and educational grading by providing detailed and immediate feedback. The study emphasizes the importance of continued refinement of AI-based evaluation systems and advocates for their responsible integration into language testing practices alongside human judgment.
Biography
This research project is a collaborative effort among three emerging scholars with a shared passion for innovation in English language teaching and learning in Vietnam. Phuong Linh Dong and Thu Giang Dang are currently English teachers at Ylang Academy, a language center based in Haiphong City. With over five years of classroom experience, they have taught a wide range of EFL learners and are particularly interested in applying modern teaching approaches and digital tools to enhance learner engagement and outcomes. Both are actively involved in professional development and have contributed to teacher training workshops and curriculum design.
Viet Anh Nguyen is an undergraduate student majoring in English Language Teaching at Ho Chi Minh City University of Education. He brings an academic lens to the group’s work, especially in areas related to student-centered methodologies.
The group’s collective research interests include technologies and AI-powered tools in language education, task-based language teaching, and intercultural competence in language teaching and learning. Their collaboration reflects a shared commitment to exploring how innovation can support more inclusive and effective English teaching practices.
