AI Predicting Lesson Plans: The Influence of Teacher and Textbook (Stage 2, 12:08 PM)

• In Event: e-Lightning Ed-Talk Session 17 - Stage 2

Abstract

In order to learn about AI’s capacity to support teacher lesson planning, this study investigates whether ChatGPT-o3 can predict how two high school mathematics teachers with different instructional orientations design lessons when given curriculum materials with different design intentions. When provided with lesson plans designed by the same teachers with other curriculum materials and an interview that captured each teacher’s approaches to instruction, the AI-generated lesson plans closely resembled the teachers’ plans in how new and old content were introduced and how activities were sequenced. When the curriculum materials and the teachers’ orientations were misaligned, the AI responded by introducing new content in a manner aligned with the teacher’s orientation. Interestingly, in all cases, AI increased the proportion of student-driven content.

Authors

• Maya Bialik, Boston University

• Pelin Jackson, Boston University

• Leslie Dietiker, Boston University