Examining the effect of AI-powered virtual-human training on STEM majors’ self-regulated learning behavior

Abstract

Read online

IntroductionStudents pursuing science, technology, engineering, and math (STEM) majors often struggle with essential skills critical to their academic success and future careers. Traditional self-regulated learning (SRL) training programs, while effective, require significant time investments from both students and instructors, limiting their feasibility in large lecture-based STEM courses.MethodsThis study investigates whether completion of three AI-powered virtual-human training modules—focused on planning, self-monitoring, and reflection—leads to increased use of corresponding MS Planner tools among STEM majors compared to a control group.ResultsResults indicate that students who did not complete the first two training modules were less likely to use MS Planner features for planning and self-monitoring; however, the reflection module did not yield comparable results.DiscussionThese findings highlight the potential of AI-powered virtual-human training as a scalable solution to enhance desirable learning behaviors among STEM majors, particularly in large and diverse classrooms. This research contributes to the understanding of effective interventions for fostering SRL behaviors in STEM education and suggests avenues for future refinement and implementation of digital training tools.

Keywords

• AI
• self-regulated learning
• STEM education
• skill training
• cognitive strategies
• virtual humans