Classification of Prompts to Generative AI and Relationship to Structural Understanding in Concept Map-Based History Learning

Abstract

In this study, we classified the prompts that students entered to receive feedback from generative AI whilecreating concept maps with its support in a social studies history lesson. Next, we divided the student ’s structuralunderstanding of inter-event relationships within the learning unit into four levels and applied our classification itemsto examine the relationship between each level and the occurrence rates of these prompt categories. Consequently, wefound differences in the occurrence rates of prompt categories between students who had a structural understanding ofevent relationships and those who did not.These findings suggest that prompt patterns can indicate students’structuralunderstanding and inform tailored feedback.