Exploring Pre-service Elementary Teachers’ Prompting as Shaping Practices When Using ChatGPT for Fraction Division Lesson Planning

This study explores how pre-service elementary teachers used ChatGPT during fraction division lesson planning by analyzing their prompts as shaping practices. A total of 1,204 prompts produced by 99 pre-service teachers during the design of Grade 6 mathematics lessons were analyzed, with individual prompts serving as units of analysis. The findings showed that prompts most frequently focused on leveraging participant structures(structuring instructional flow and alinging with curricular materials), and framing task constraints, indicating that ChatGPT was primarily used to support the practical construction and refinement of lesson plans. Prompts frequently included textbook materials or lesson plan references when requesting revisions and elaborations, suggesting efforts to obtain more contextually appropriate instructional outputs. Iterative refinement was common, as participants repeatedly revised and evaluated AI-generated responses rather than accepting them uncritically. In contrast, prompts explicitly addressing mathematical meaning or students’ reasoning appeared relatively limited. The findings suggest that generative AI primarily functioned as a mediating tool for organizing and refining lesson-ready materials while also revealing pre-service teachers’ instructional priorities during lesson planning.