The Impact of AI Chatbots as Digital Speaking Partners on EFL Learners’ Speaking Proficiency: A Quantitative Study

Abstract

The use of artificial intelligence (AI) technologies is now widely integrated into language education, providing effective ways to assist learners' speaking practice in EFL classrooms. This research explores the possibility of applying AI chatbots as digital speaking partners to improve students’ speaking competence. A quantitative quasi-experimental design was used with 60 third-semester university students divided into an experimental and a control group. The study's experimental group used an AI-based chatbot to practice speaking, while the control group received traditional instruction. The students’ speaking performance was measured by using an analytic rubric that consisted of pronunciation, grammar, vocabulary, fluency, and comprehension. The experimental group’s mean score on the pre-test was 80.50, and on the post-test, it increased to 85.07. In contrast, the control group had only improved from 79.47 to merely 79.83 on the two tests, respectively. The data was analyzed using an independent samples t-test, confirming a significant difference between the two groups (t = 3.241, p = 0.002), indicating that students who practiced speaking with AI chatbots demonstrated significantly greater improvement in their speaking performance compared to those receiving traditional instruction. These results reflect that utilizing AI chatbots in speaking practice has the potential to significantly improve EFL learners’ speaking proficiency by providing more opportunities for interactive practice.