Exploring the Efficacy of ChatGPT in Language Teaching

Abstract

This study examines the potential of using Artificial Intelligence (AI), specifically ChatGPT, for language teaching. As AI technology continues to advance and be integrated into education, there is growing interest in its role in improving language learning and evaluation. To explore the efficacy of ChatGPT, it was assigned to evaluate ten writing essays from advanced English students, providing detailed feedback and assigning grades based on the established criteria. Then, the results were viewed by an experienced senior teacher who specializes in teaching advanced English writing. The study employs a qualitative approach, with an in-depth interview with the teacher to explore her perspective on the reliability and effectiveness of ChatGPT's evaluation. The findings show a significant similarity between the grades assigned by the teacher and ChatGPT for the set of writing essays. The results from interviewing the teachers showed that ChatGPT is a useful tool for reducing teachers' workload but can only replace human teachers partially. This study supports AI as a potential complement to human teaching and underscores the importance of human expertise for comprehensive feedback, suggesting AI-human collaboration for improved writing instruction and evaluation.