GenAI as a Design Partner: Building Preservice STEM Teachers’ Phenomenon-Based Lesson Design and AI Literacy

Abstract

This one-year pilot study examines how preservice secondary STEM teachers use generative AI (GenAI) to design phenomenon-based science lessons, and how their AI literacy develops through that work. Texas now requires phenomenon-based instruction, but preservice teachers struggle to select a suitable anchoring phenomenon for a given standard and to anticipate the range of student ideas and investigations. Six UTeach preservice teachers at the University of Texas Rio Grande Valley will complete a five-day workshop in two cohorts of three. The first two days immerse participants in phenomenon-based learning. The last three days guide them through an AI-supported lesson design cycle using a staged prompt framework that returns multiple options for participants to compare, justify, and revise. Participants complete an independent design task before the workshop and again three to four weeks after it. Using a multiple-case study design, the project collects AI literacy surveys, interviews, workshop video recordings, lesson designs, GenAI prompt logs, and reflective journals. The findings will show how AI literacy and phenomenon-based design expertise develop together. They will also provide preliminary data and tested instruments for a larger NSF proposal.