Supporting Responsive Teaching with GenAI-Powered Formative Assessment (Stage 2, 9:50 AM)

• In Event: e-Lightning Ed-Talk Session 16 - Stage 2

Abstract

We explored teachers’ reactions to two tools for responsive teaching, created with generative artificial intelligence (GenAI): (1) individualized student feedback, and (2) teacher-facing visualizations displaying formative assessment results. Both tools incorporated a multimodal large language model’s analysis of students’ drawn models and written explanations from a science unit on energy transformation. Thematic analysis of teacher interviews revealed three key themes. First, teachers wanted personalized feedback that promoted targeted follow-up. Second, they desired contextualized feedback that aligned with instructional goals, classroom contexts, and teachers’ customization. Third, they had concerns about transparency regarding the feedback generation process and ethics. The research contributes to emerging understanding of how GenAI can support K–12 science education, with a focus on enhancing formative feedback practices.

Authors

• Idris Solola, Utah State University

• Ravi Sinha, Utah State University

• Hillary Lucille Swanson, Utah State University

• Rida Munir, Utah State University

• Ha Nguyen, University of North Carolina - Chapel Hill