Preservice Teacher AI-Supported Collaborative Noticing for Asset-Based Mathematics Pedagogy

Abstract This study aims to unpack the learning mechanisms that undergird the use of AI simulations in teacher education, illuminating how preservice teachers learn to notice students’ mathematical thinking through simulated interactions and collaborative sensemaking. This exploratory case study introduces five AI agents designed to simulate elementary students and support preservice teachers in developing noticing skills for asset-based pedagogy. The preservice teachers worked in pairs to practice conducting diagnostic interviews with five AI agents. The AI agents vary in their fraction thinking and mathematical representations. Findings show that preservice teachers used various noticing discourse strategies to make sense of AI agents’ mathematical representations, demonstrating an asset-based mindset. Through these interactions, they attended to students’ mathematical strengths, interpreted their thinking, and made informed pedagogical decisions. Preservice teachers also reported that peer discussion further supported their ability to elicit and interpret AI students’ mathematical thinking. Our work demonstrates the intentional integration of carefully designed AI agents within AI-supported collaborative learning environments can enhance preservice teachers’ noticing skills.