Comparative Analysis of AI and Human Instructors in Providing Written Corrective Feedback: A Case Study on ChatGPT

Abstract

Corrective feedback is one of the instructional methods utilised in learning writing. Any indication that a student’s expression might contain an error of some sort is referred to as corrective feedback. Obtaining feedback is a component of the process of writing improvement. Traditionally, this feedback has been provided by human instructors, who meticulously review students' work to pinpoint areas of improvement. However, with the advent of AI technology, particularly exemplified by the ChatGPT model, the landscape of feedback provision is undergoing a significant transformation. Thus, this study aims to investigate the differences between AI and human instructors in providing written corrective feedback, focusing on the ChatGPT AI model within the context of ESL writing instruction in a higher learning institution in Malaysia. The study addresses a gap in research by examining the types of corrective feedback offered by ESL instructors and ChatGPT, aiming to uncover potential differences between them. The research objectives include identifying the types of corrective feedback provided by ESL instructors and ChatGPT and exploring any disparities between them. The study's significance lies in its contribution to understanding written corrective feedback in ESL writing, offering insights for researchers, educators, learners, and policymakers. This research seeks to inform pedagogical practices and educational policies by bridging traditional ESL instruction with emerging AI technologies. Overall, this study aims to shed light on the potential of AI-driven tools like ChatGPT to enhance ESL writing instruction while recognising the enduring value of human expertise in language teaching.