Can ChatGPT Replace English Teachers? A Comparative Study of Feedback Quality and Learner Perception

This study investigates the efficacy of ChatGPT-generated feedback compared to teacher feedback in English as a Foreign Language (EFL) writing contexts, examining both feedback quality dimensions and learner perceptions. Drawing on recent comparative research, the study synthesizes findings from multiple empirical investigations examining epistemic strategies, delivery methods, emotional engagement, and learner trust. Analysis reveals that while ChatGPT provides immediate, detailed, and grammatically sound feedback with ۶۸% overlap with instructor comments on similar writing features, teacher feedback demonstrates greater assertiveness, dialogic engagement, contextual appropriateness, and emotional resonance. Teacher feedback significantly enhances Foreign Language Enjoyment and builds higher learner trust, leading to more sustained writing improvements on delayed post-tests. Learners consistently express preference for teacher feedback despite acknowledging ChatGPT's utility for immediate support. The findings suggest that optimal feedback practices require blended models integrating AI efficiency with human instructors' interpersonal depth and adaptive expertise, rather than viewing ChatGPT as a potential replacement for English teachers. Implications for pedagogical practice and AI tool development are discussed.

کلیدواژه ها:

ChatGPT ، teacher feedback ، EFL writing ، learner perception ، artificial intelligence in education ، comparative study