#282: Teacher Perceptions of AI-Generated Feedback on Student Scientific Models

Effective feedback is essential for supporting student learning in complex tasks such as scientific modeling, yet teachers often struggle to provide high-quality feedback. This study explored the utility of MLLM-generated feedback from teachers’ perspectives through a focus group with four middle school science teachers. Overall, participants regarded the feedback as promising. They valued the feedback’s degree of detail, supportive tone, and potential to deepen teachers’ knowledge and enrich classroom discussions. Participants also raised important concerns about the feedback’s alignment with instructional contexts, length, and readability.