The Impact of ChatGPT’s Immediate Feedbackon EFL Learners' Writing Accuracy versusAppropriateness

This study investigated the impact of OPenAI’s ChatGPT-۳.۵ immediate feedback on the writing improvement of Englishas a foreign language (EFL) learners’ writing accuracy versus appropriateness. To this end, ۶۰ upper-intermediate GeneralIELTS candidates from Iran University of Medical Sciences’ Language Department were recruited based on convenientsampling. After checking for their homogeneity, they were divided into control and experimental groups. The studyemployed a quasi-experimental pre-test, treatment and post-test design. The experimental group was treated differently onassignment feedback it received via Artificial Intelligence (AI). Results of independent-samples t-tests, paired-samples ttests,and MANOVA analyses revealed positive impacts of ChatGPT’s instantaneous feedback on the participants’ writingaccuracy and appropriateness. The findings suggest that AI immediate feedback can significantly enhance EFL writingperformance while considering the importance of human intervention and guidance. However, the study also identifiedchallenges related to overreliance on AI tools and the limitations in capturing nuanced language aspects. The implicationsof this research contribute to the understanding of how AI technology can be utilized to improve EFL writing instruction,informing pedagogical practices and highlighting the need for a balanced approach that combines AI tools with humansupport

Keywords:

Artificial Intelligence (AI) , Immediate Feedback , ChatGPT , Writing Accuracy and Appropriateness