Chatbot-Supported Language Learning: Effects on Pronunciation and Vocabulary Acquisition Among Novice Thai Learners

This study investigates the effectiveness of a chatbot-based learning system for beginner Thai language acquisition, with a focus on pronunciation accuracy, immediate vocabulary learning outcomes, and learner experience. Grounded in the Computer-Assisted Language Learning framework, the proposed system integrates real-time pronunciation feedback, multimodal input, and gamified learning activities within a LINE-based chatbot environment. A mixed-methods study incorporating a randomized controlled pretest-posttest design was conducted with 80 Chinese-speaking participants, who were randomly assigned to either a chatbot-assisted learning group or a self-study control group over an 8-week intervention period. Quantitative results from pre- and immediate posttests indicate that the chatbot group achieved significantly greater improvements in pronunciation accuracy and vocabulary acquisition than the control group. Because no delayed retention assessment was conducted, the findings should be interpreted as evidence of immediate learning gains rather than evidence of sustained vocabulary retention. Usability and user experience were evaluated using the System Usability Scale and the User Experience Questionnaire, with findings demonstrating high levels of usability, engagement, and learner satisfaction. Qualitative feedback from chatbot-group participants further highlighted the role of real-time feedback and interactive design in enhancing motivation and confidence. The findings suggest that this integrated LINE-based chatbot was associated with greater immediate pronunciation and vocabulary gains than static self-study materials under the conditions of this study. This study contributes to the advancement of chatbot-assisted language learning by providing empirical evidence and design implications for developing interactive and learner-centered systems in less commonly taught languages.