Presentation Enhancing EFL Learning: Generative AI vs. Human Interactions in Willingness to Communicate and Grit more

Advancements in generative artificial intelligence (GenAI) have opened new avenues for interactive, adaptive, and personalized language learning experiences, yet limited research has compared the effects of GenAI chatbots and human interlocutors on elementary English as a Foreign Language (EFL) learners’ willingness to communicate (WTC) and grit. In this presentation, I will share findings from a 12-week study exploring how GenAI chatbots compare with human interlocutors in enhancing elementary EFL learners’ WTC and grit. Fifty-seven third-grade students in Taiwan participated in this 12-week study. Participants engaged in 10-minute communication activities during their English class weekly and were divided into two groups: the Bot Group (N=30), who interacted with CoolE Bot, and the No-Bot Group (N=27), who participated in peer interactions. The results demonstrated that CoolE Bot significantly improved learners’ WTC and grit compared to peer interactions. Qualitative analysis highlighted five key benefits of CoolE Bot: (1) providing authentic communication opportunities, (2) facilitating dynamic and coherent interactions, (3) offering contextually appropriate and personalized responses, (4) adopting multi-interactive roles, and (5) fostering low-pressure environments with personalized feedback. These features enhanced participants’ WTC and grit. The findings suggest that GenAI chatbots can complement conventional human-mediated instruction by providing flexible and personalized language practice.