Teachers’ Use of Generative Artificial Intelligence for Designing Science Lessons in Support of Environmental Science Agency

Abstract Generative Artificial Intelligence (GenAI) holds promise as a tool to assist teachers in addressing specific instructional design and implementation tasks. This study explores how teachers in a master’s-level elementary science methods course used Large Language Model AI, as one type of GenAI, to design lessons fostering students’ environmental science agency (ESA). The course introduced ESA as one of pedagogical constructs that emphasize the importance of science learning relevant and consequential to students’ lives, and encouraged teachers to use GenAI for their ESA-focused lesson design projects. Data sources included surveys, lesson design documents, GenAI chat logs, and field notes. Findings, derived through constant comparative analysis, highlight GenAI’s utility in three areas: (1) deepening teachers’ conceptual understanding; (2) developing detailed content of lesson components; and (3) integrating ESA into science instruction. Teachers also reported the challenges they encountered and addressed mainly in two ways: (1) ensuring the quality of GenAI-generated responses and (2) aligning these responses with their specific lesson visions. These findings were discussed in terms of teacher perspectives on ESA integration and their effort to take authorship and responsibility in their project process and outcomes. The study concludes by calling for follow-up teacher education practice and research focused on creating science learning opportunities, including those with proper and ethical use of GenAI, that go beyond knowledge enhancement to fostering civic participation aimed at justice-oriented and transformative changes, as envisioned by the concept and practices of ESA.