Leveraging ChatGPT’s advanced voice mode for enhancing oral-aural skills in higher education language learning: A mixed-methods study

Texto completo

Examines the effects of ChatGPT's Advanced Voice Mode on the development of university students' oral proficiency. Following a sequential explanatory mixed-methods design, the study measured improvement over an 8-week intervention and then gathered learners' perspectives through semi-structured interviews. Quantitative findings showed that fluency and lexical resources increased significantly, but pronunciation did not. The qualitative data focus on three themes: the potential of AI for low-anxiety practice, learners' frustration with the lack of nuanced feedback on accent, and difficulties with real-time corrective functions. The authors propose a «Mixed Dialogical Model» combining technology-mediated practice and expert human tutoring.