Paper Summary
Share...

Direct link:

Explore Teacher Talk Moves in a Generative AI-Powered Virtual Reality Classroom Dialog Simulation

Wed, April 23, 10:50am to 12:20pm MDT (10:50am to 12:20pm MDT), The Colorado Convention Center, Floor: Meeting Room Level, Room 710

Abstract

Productive classroom dialogs between teachers and students are crucial for educational success. Large language models (LLM) offer opportunities for simulating classroom dialogs with the ability to facilitate near-authentic social interactions. This study pioneered at exploring preservice teachers' talk moves when teaching generative AI-powered conversational student agents in a virtual reality (VR) classroom. Twenty-three preservice teachers participated in a 10-minute teaching simulation aimed at developing students' argumentation skills. The study measured the number and length of teacher and student utterances, as well as the types of talk moves. Initial findings indicated a higher frequency of teacher utterances but longer student responses. This research underscores the high potential of generative AI for training dialogic pedagogy in promoting students' argumentation skills.

Authors