We find ourselves at a pivotal moment where the role of generative AI and other technological tools is reshaping the way we teach and learn languages. The question that lies before us is not whether these innovations will shape the future, but rather how we, as educators, will harness their potential to create meaningful and effective language instruction.
The title of our conference, No Fate: The Future is Not Set, underscores our belief that the future is not predetermined. We hope that our conference will serve both as a platform for collaboration and a catalyst for change as it is via the collective effort of educators, researchers, and innovators that the trajectory of language teaching and learning will be determined. By fostering collaboration, sharing insights, and pushing the boundaries of what is possible, we can shape the future of language education.
This study aims to investigate the impact of AI-integrated blended learning on foreign language speaking anxiety (FLSA) among Chinese IELTS candidates. This study was conducted at New Channel, an overseas study education institution in China. 30 learners and an IELTS speaking instructor participated in this four-week AI-integrated blended learning period. Data were collected using the IELTS Speaking Anxiety Scale (ISAS) survey (n=30) and mock IELTS speaking tests (n=30) to compare pre-and post-intervention anxiety levels and learning outcomes. These were further supplemented by semi-structured interviews (n=6) and teacher feedback to gather learners’ in-class performance and their perspectives of using AI in IELTS preparation. Quantitative data were employed SPSS 27.0 to analyze the statistics from 30 sets of ISAS surveys and mock IELTS speaking test before and after intervention, while qualitative data were analyzed using thematic analysis to examine the teacher feedback forum and interview transcripts from six interviewees. The results from the ISAS survey indicated that Chinese IELTS candidates’ speaking anxiety were significantly decreased (MD=1.59) after AI-integrated blended learning, particularly in anxiety aspects related to fluency and coherence, grammatical and lexical rules, and IELTS test preparation. Additionally, there was a notable improvement in IELTS speaking scores (MD=0.42) after AI intervention, particularly in the Lexical Resource (LR) dimension (MD=0.7). Teacher feedback and interviews further expressed positive attitudes towards this AI application’s immediate feedback and realistic exam simulation. However, the results also revealed the limitations of AI in reducing the psychological and physiological aspects of anxiety that students experience in high-pressure exam environments. This study demonstrates how AI-integrated blended learning could be applied to impact learners’ speaking anxiety and oral performance in the IELTS examination context. The findings also offer insights into further research and development in AI tools aimed at reducing speaking anxiety and facilitating speaking practice in high-stakes environments.
