How to Improve Pre-service Teachers' Classroom Organization Ability?——Design and Application Experiment of a Multi-Role Scenario Simulation Agent

Currently, generative artificial intelligence technology is profoundly reshaping the educational form, and the training of pre-service teachers' teaching ability is gradually shifting from a traditional training model centered on “interpersonal interaction” to an intelligent training model of “human-machine integration”. Against this background, focusing on the issue of “how to improve pre-service teachers' classroom organization ability”, this paper designed a multi-role scenario simulation agent relied on Gagne's instructional design theory and Tomlinson's differentiated instruction theory, and combined the need to improve pre-service teachers' classroom organization ability. It was pointed out in this paper that the agent was centered on teaching scenario simulation, student role design, student behavior simulation, guided training mechanisms and response evaluation, with an emphasis on providing the classroom organization ability training and feedback support for pre-service teachers. Subsequently, an application experiment was conducted on an intelligent teaching skill training platform for pre-service teachers to verify the effectiveness of the multi-role scenario simulation agent in enhancing pre-service teachers' classroom management skills. The results showed that the application of the multi-role scenario simulation agent can help improve pre-service teachers' classroom organization ability and relieves their pressure in dealing with classroom organization problems. Meanwhile, pre-service teachers achieved remarkable improvement in designing classroom organization strategies in specific tasks. The multi-role scenario simulation agent designed in this paper and its application not only offered practical references for the refined design of educational agents driven by teaching theories, but also provided empirical evidence for the empowerment of teacher education through multi-role scenario simulation agents.