AI-generated vs teacher feedback: A comparative study on EFL learners’ public speaking anxiety and oral presentation performance

Abstract

This comparative study investigates the impact of artificial intelligence (AI)-generated feedback versus teacher-provided feedback on public speaking anxiety (PSA) and oral presentation performance among Turkish EFL learners. Adopting a quasi-experimental pre-test–post-test control group research design, thirty university students took part in the study. They were assigned to either an experimental group receiving AI feedback or a control group receiving teacher feedback. Over five weeks, participants delivered presentations and received feedback based on their group assignment. Quantitative data were collected through the Public Speaking Anxiety Scale (PSAS) to measure anxiety levels and a standardized rubric to evaluate oral presentation performance. The findings showed that both feedback types significantly decreased students’ public speaking anxiety and improved their presentation skills. Specifically, AI-generated feedback resulted in a notably greater reduction in PSA, whereas both groups achieved similar improvements in oral performance. Qualitative data from interviews with AI group participants supported the quantitative findings, underlining increased self-awareness, practical guidance, and enhanced confidence. The study concludes that while both feedback sources are effective for skill development, AI-generated feedback holds a distinct advantage in decreasing public speaking anxiety. This underlines the value of AI as a complementary tool in EFL speaking instruction, supporting a hybrid pedagogical model that harnesses the strengths of both automated and human feedback.