Strategic genAI Use in Proving: Effects of Sequence on Student Learning and Reasoning

• In Event: AERA Poster Session Saturday 7:45 am
  In Poster Session: From Adoption to Transformation: AI Integration, Critical Pedagogy, and Sustaining Futures in Teacher Preparation

Abstract

This study examines how the timing of generative artificial intelligence (genAI) use affects undergraduate students' construction of mathematical proofs. We compare two instructional sequences—Initial Independent Proving (IIP) and Initial Dependent Proving (IDP)—to explore how students engage with genAI-generated proofs and reflect on their own reasoning. Findings indicate that students who first attempt problems independently revise more deeply and transfer ideas more flexibly, while early genAI use offers efficiency but may hinder conceptual engagement. These insights inform instructional design principles for integrating genAI in mathematics education.

Authors

• Hyunkyoung Yoon, California State Polytechnic University, Pomona

• Yujin Lee, Kangwon National University

• Kyungwon Lee, Seoul National University

• Oh Nam Kwon, Seoul National University