Beyond Theory: A Mixed-Methods Investigation of Postgraduate Engagement with ChatGPT for IELTS Speaking

Abstract This study investigates the burgeoning interest in ChatGPT's role in education, with a particular emphasis on its effects on postgraduate students' English-speaking proficiency, a topic previously explored mainly through theoretical lenses with scant empirical evidence. The research aims to fill this gap by examining the experiences and perceptions of Chinese postgraduate students using ChatGPT for IELTS speaking practice. Employing a mixed-methods approach, the study combines experimental engagement with ChatGPT, analysis of reflective diaries, and responses from questionnaires to gauge the platform's efficacy in fostering enthusiasm for English language practice. The findings indicate a positive impact of ChatGPT on learners' engagement, as evidenced by an increase in practice frequency. A significant correlation was observed between regular interaction with ChatGPT and improvements in English grammar skills among participants. Thematic analysis of reflective diaries highlighted ChatGPT's contribution to language learning and students' proactive approach to acquiring English-speaking skills. Nonetheless, there was a consensus among students for a need for more human-like interactions from ChatGPT. The study contributes valuable insights into the potential of chatbots to enhance English-speaking skills in EFL contexts, offering practical recommendations for educators and developers to optimize chatbot learning experiences.