AI AS THE STUDENT: CONCEPTUAL DESIGN OF A SIMULATION ENVIRONMENT FOR DEVELOPING TEACHER DISCOURSE COMPETENCE

Abstract

This paper presents the conceptual design of a simulation-based training system that supports pre-service teachers in developing discourse competence for facilitating accountable classroom talk on socio-scientific issues (SSI). The system is grounded in research on dialogic teaching, conceptual change, teacher reflection, and the TPACK framework. It comprises five interconnected modules: knowledge training, simulation preparation, dialogic simulation, discourse analytics, and reflective feedback. The components are powered by generative AI based on large language models (LLMs), which provide adaptive support across the entire learning cycle from explaining pedagogical concepts to generating realistic student responses, analyzing teaching moves, and delivering personalized feedback. The system creates a practice-oriented, low-risk environment in which pre-service teachers can experiment with discourse strategies, manage epistemic and emotional complexity, and reflect on their instructional decisions. In addition to demonstrating the pedagogical potential of generative AI, design also raises broader questions about the development of trustworthy educational technologies. Future work includes empirical evaluation in the form of a pilot study and iterative refinement through design-based research.