The Use of Customized GenAI Chatbots to Support Preservice Teachers’ Understanding of Collective Mathematical Argumentation

• In Event: AERA Poster Session Sunday 11:45 am
  In Poster Session: SIG-RME Poster Session: Explorations of the Mathematics Teaching and Learning Experience

Abstract

This study introduces the GRASP (Generative Role-Play AI Simulation for Pedagogy) model, which employs customized ChatGPT-based agents to support preservice mathematics teachers (PMTs) in a four-phase practice-based learning cycle: initiating, exploring, enacting, and reflecting. Focusing on the first two phases, we investigated how GRASP supports PMTs’ understanding of collective mathematical argumentation. Ten PMTs enrolled in a mathematics education course participated in the study. Analysis of their assignments and reflections revealed that most demonstrated a basic to advanced understanding of collective argumentation. Reflections further indicated that interactions with AI chatbots facilitated multi-layered learning. Findings highlight the potential of GenAI in mathematics teacher education and point to future directions for leveraging AI to support PMTs’ development of pedagogical practices.

Authors

• Yuling Zhuang, Texas A&M University

• Xiangquan Yao, Pennsylvania State University

• Trina J. Davis, Texas A&M University

• Wisdom Yao Nudze, Texas A&M University