#1097: Self-Regulated Learning with Generative AI: Student Strategies, Performance-Based Profiles, and Instructional Implications

This mixed-methods study examines how students enact self-regulated learning (SRL) with GenAI in an authentic undergraduate chemical engineering course. Using an activity-centric SRL lens, we triangulated interviews with students (n = 12) and instructors (n = 3), a post-course survey (n = 40), course-tool interaction logs, and AI-supported student artifacts. Findings show that GenAI supported SRL across forethought, performance, and reflection by helping students initiate tasks, refine ideas through dialogue, verify outputs, and review their work. Across performance groups, three recurrent profiles emerged: Prudent Architects, Struggling Explorers, and Anxious GenAI functioned primarily as a learning-intention amplifier in SRL, strengthening deeper regulation when mastery goals were strong and reinforcing shortcut-oriented strategies when performance goals dominated. We discuss targeted instructional scaffolds for fostering critical and ethical AI-supported learning.