The impact of chatbot-based speaking practice on learners' foreign language speaking anxiety and motivation

Foreign language speaking anxiety is considered as a significant affective barrier that negatively impacts students' oral performance and speaking motivation in the target language. However, the effects of AI-supported applications on these variables remain relatively limited in the context of foreign language education. This study aims to examine the effects of AI-supported spoken language applications on English language teaching students' foreign language speaking anxiety and second language speaking motivation. The research was conducted using a quasi-experimental, single-group pretest–posttest design and adopted a mixed-methods approach. Seventy-five freshman students enrolled in the English teaching program at a state university in Turkiye during the 2024–2025 academic year participated in the study. Data were collected through the Foreign Language Speaking Anxiety Questionnaire and the L2 Speaking Motivation Scale, as well as participant portfolios, post-intervention student reflections, and researcher diaries; quantitative analyses and thematic qualitative analyses were used together for evaluation. The findings revealed that while there was no statistically significant change in scale scores, qualitative data indicated positive developments in students' speaking anxiety and motivation level. Consequently, it was observed that AI-supported speaking applications are more effective as a complementary pedagogical tool that supports the process when integrated with real-life and classroom interactions, rather than as a sufficient intervention on their own.