THE ROLE OF GENERATIVE AI IN DEVELOPING IELTS WRITING SKILLS

Abstract

This exploratory quasi-experimental study compared generative-AI-assisted feedback with teacher-led feedback in IELTS Writing Task 2 practice. Twenty pre-intermediate learners aged 16–18 from one private language school were assigned to two equal groups. Both groups completed a pre-test, a four-week intervention comprising 16 writing activities, and a post-test under comparable 40-minute conditions. The AI-assisted group used ChatGPT-generated comments and model revisions, whereas the teacher-led group received feedback from an instructor on the same broad dimensions: task response, coherence and cohesion, lexical resource, and grammatical range and accuracy. At post-test, the AI-assisted group obtained a mean overall band score of 6.5 and the teacher-led group 7.0. Classroom observations suggested that AI support was especially useful for idea generation and lexical alternatives, while teacher feedback was more consistently acted upon and appeared stronger for grammar and essay organization. Because individual pre-/post-test scores, dispersion measures, and validated perception data were unavailable, the study does not claim statistical significance or causal superiority. The findings support a pedagogically supervised, hybrid use of generative AI rather than its use as a substitute for teacher judgement.