Students’ Perceptions of Feedback on Fluency and Pronunciation from AI and Teachers

Feedback on students' speaking performance is crucial in teaching and learning contexts. However, since human teachers may not always be available, artificial intelligence (AI) technology can serve as an alternative for providing feedback. This study examined students' perceptions of feedback on fluency and pronunciation provided by both a human tutor and AI. The research involved seven L1 Indonesian participants enrolled in an online English course. Feedback was based on a speaking task about college admission, and their audio recordings were analyzed for fluency and pronunciation. Semi-structured interviews were conducted and analyzed using thematic analysis to explore participants' perceptions of feedback from AI and human teachers. The results indicated that most students appreciated feedback from both sources but preferred AI for fluency feedback and human tutors for pronunciation feedback. As a result, participants generally favored a hybrid approach more comprehensive feedback, as students perceived that each source had strengths that complemented the other's weaknesses.