Developing an Artificial Intelligence Simulator to Support Mathematics Teacher Questioning: The A.I. Classroom Teaching Simulator (ACTS) (Poster 2)

• In Event: AERA Poster Session Sixteen
  In Poster Session: Emerging Technology

Abstract

This paper reports on the development of a scalable, low-cost teaching simulator to support mathematics teacher questioning practices. Teachers use the AI-based Classroom Teaching Simulator (ACTS) by having voice-based natural language interactions with a virtual student in an immersive 3D setting. To demonstrate its effectiveness, we conducted a formative evaluation of an elementary-level mathematical scenario with expert teachers and examined their perceptions of the simulator experience and the natural language interactions with the virtual student. Perceived affordances of the system included having the mathematical task represented in the environment, but constraints included the inability to interact with the representation as well as opportunities to develop rapport with the virtual student. Results provide insights for designers of similar simulated practice environments.

Authors

• Jim Bywater, James Madison University
  Presenting Author

• Jennie Chiu, University of Virginia
  Presenting Author

• Sodiq Adewole, The University of Virginia
  Non-Presenting Author

• Ginger S. Watson, Old Dominion University
  Non-Presenting Author

• Don Brown, University of Virginia
  Non-Presenting Author