Presentation AI-Supported Contextualised Scenario Practice in Tourism English: Learning Outcomes and Learner Feedback in an ESP Course more

Technology-enhanced education enables more interactive, personalised language learning. This project examined how GenAI (e.g., ChatGPT-4o) supported university students without a tourism background in an ESP Tourism English course. Although academically strong, they were novice to tourism discourse and workplace routines, creating a mismatch between course expectations (service-oriented interactional competence) and prior textbook-driven learning with limited oral rehearsal. This gap lowered perceived relevance, increased speaking anxiety, and led to uneven role-play participation. Therefore, the intervention targeted motivation, learner autonomy, and listening–speaking performance for tourism-service encounters. Over 18 weeks, students completed AI-supported brainstorming, itinerary planning, collaborative tasks, and scenario-based conversational practice in workplace simulations, producing recordings and task outputs. The design provided scalable, repeatable interaction practice with immediate language support beyond what teacher feedback alone can offer in limited class time. Qualitative triangulation of student artefacts, recordings, AI dialogue logs, and reflections indicated improvements in four areas. Students demonstrated stronger oral expression and situational responsiveness, greater appropriacy in service interactions, higher motivation due to reduced speaking anxiety, and improved AI literacy for revising language and organising information. Student feedback also noted vocabulary gaps and the need for clearer AI-use guidance and a better balance between AI practice and human interaction.