Presentation Using AI-Mediated Video and Chatbot Tasks to Scaffold Speaking in CLIL: A Pilot Study more

This pilot study explores how technology-mediated speaking tasks can support the development of oral summarization in a CLIL (Content and Language Integrated Learning) course at a Japanese university. Using EnglishCentral as a digital platform, first-year EFL students engaged with video content and completed repeated oral summarization tasks over the course of a semester. In addition to video viewing and vocabulary-learning activities, learners interacted with MIMI, EnglishCentral’s AI discussion chatbot, to rehearse ideas and language prior to recording spoken summaries. The project investigates how the combination of authentic video input, AI-mediated interaction, and repeated recording tasks may support students’ emerging summarization skills. Particular attention is given to learners’ use of content-specific vocabulary, complexity of spoken language, and overall comprehensibility. As this study represents an initial investigation, qualitative analysis centers on six students who completed all required summary recordings. Audio recordings and platform data are examined to identify patterns of development across time. While large-scale statistical claims are not the goal of this phase, the study aims to demonstrate how CALL tools can scaffold spoken output in cognitively demanding CLIL contexts. Findings will inform the design of a subsequent larger-scale study investigating longitudinal development of oral summarization through AI-supported CLIL tasks.