Phased Evolution of Teacher-AI Collaboration for the Effective Mentoring in Teacher Education

This study explores the phased evolution of teacherAI collaboration (TAC) to improve mentoring effectiveness in teacher education. Adopting an action research approach, TAC was implemented across two cycles, engaging 42 fourth-year university students in interview practice sessions, alternating between providing mentoring by ChatGPT-4 and using human mentors. Stress levels, communication efficacy, and feedback personalization were assessed using quantitative and qualitative surveys. The findings indicate that AI-assisted mentoring reduced mentors' workload and provided a structured platform for individualized feedback. The students reported reduced stress and greater ease when interacting with the AI. The results of this study highlight the potential of TAC to enhance educational support in resource-constrained settings, emphasizing the need for prompt engineering to further refine the AI feedback quality.