THE IMPACT OF GENERATIVE AI ON EFL ACADEMIC WRITING SKILLS: A QUASI-EXPERIMENTAL STUDY IN UZBEKISTAN

This quasi-experimental study examines the impact of generative artificial intelligence (AI) on EFL academic writing skills at Fergana State University, Uzbekistan. Grounded in Vygotsky’s Zone of Proximal Development, Self-Regulated Learning theory, and the Process Writing Approach, the study investigates whether guided AI-assisted instruction improves writing performance relative to traditional teacher-led instruction. Sixty intermediate-level students (B1–B2 CEFR) completed a seven-week intervention, with pre- and post-test argumentative essays assessed using an IELTS-based analytic rubric. Quantitative data are analyzed using paired and independent-samples t-tests; qualitative data are gathered through perception surveys and semi-structured interviews. The study specifically explores the differential effects of AI on lexical resource, grammatical accuracy, coherence, and higher-order argumentation, alongside student attitudes toward AI use and its implications for writing autonomy and academic integrity. Findings are expected to contribute empirical evidence to the limited body of quasi-experimental AI research in Central Asian EFL higher education.