AI as a Bridge Between Theory and Practice in Teacher Education: How Textual and Metaverse-Based ChatGPT Modalities Shape Help-Seeking

This study explored whether immersion mode affects the efficacy of artificial intelligence (AI) driven learning modalities in developing teaching efficacy among pre-service teachers. This study explored the potential of AI-based technologies to improve pre-service special education teachers’ self-efficacy in solving classroom challenges, such as managing student behavioral problems, that are often difficult to master through traditional college courses. This study examined the experiences of pre-service teachers (n=17) utilizing an AI expert system grounded in chat generative pre-trained transformer (ChatGPT) via two modalities: virtual reality (VR) and text-based settings. The participants were divided into three groups to examine self-efficacy regarding help-seeking behavior from (1) a traditional web search group as a comparison, (2) a text-based ChatGPT group, and (3) a VR-based voice interactive ChatGPT group, exploring different modes of immersion. The findings indicated that text-based ChatGPT was the most effective in improving pre-service teachers’ self-efficacy with managing problematic behaviors, followed by VR-based ChatGPT, while the web search group showed comparatively low efficacy. These findings highlight the importance of simplicity, cognitive efficiency, and user-friendly interfaces of text-based ChatGPT in educational technology design. This study emphasizes the potential benefits of integrating AI systems into teacher training programs, intended to connect theoretical knowledge with practical application in special education settings. This study holds considerable research value and significance, suggesting that text-based ChatGPT can be an effective medium that can help pre-service special education teachers in interpreting theoretical educational content acquired in college to actual educational environments.