The Future of Education

This presentation introduces a comparative project examining English-speaking practice with AI avatars in a virtual reality (VR) environment. The project has three objectives: (1) to measure changes in speaking proficiency, (2) to examine changes in psychological factors such as foreign language anxiety (FLA) and motivation, and (3) to investigate learners’ technology acceptance and perceived usefulness of VR-based language learning through the Technology Acceptance Model (TAM). The learning context combines AI-avatar speaking practice with intercultural interaction opportunities for Japanese and Italian students using Immerse (desktop), enabling role-play with avatars and real-time peer exchanges. To establish baseline profiles prior to VR implementation, we report pre-test questionnaire results from 52 participants (Italy n=34; Japan n=15). FLA was assessed with eight items on a 1–5 agreement scale; two positively worded items were reverse-coded so that higher scores indicate higher anxiety. Baseline FLA levels were comparable across groups (Italy M=3.19, SD=0.86; Japan M=3.27, SD=0.84), suggesting similar initial anxiety intensity. Motivation was measured on a 1–7 scale (1 “does not correspond at all” to 7 “corresponds exactly”) and analysed by subcomponents. Japanese students reported higher extrinsic regulation (Italy M=3.68; Japan M=4.92) and higher identified regulation (Italy M=5.68; Japan M=6.60), while amotivation was low and similar in both groups (Italy M=1.68; Japan M=1.82). Technology acceptance (TAM-based; 17 items on a 1–5 agreement scale; negative items reverse-coded) was positive in both groups, with a higher mean in Japan (Italy M=3.64, SD=0.57; Japan M=3.91, SD=0.42). Overall, these pre-test findings provide a baseline for interpreting subsequent VR-based learning outcomes and for modelling country-by-time patterns in anxiety, motivation, and technology acceptance.