Arif Rachmatullah, Krystal Thomas, and Jessica Mislevy
March 2026

As online STEM course enrollment continues to grow, challenges such as high dropout rates and reduced engagement persist, particularly among students from historically marginalized groups. Self-directed learning (SDL), which encompasses motivation, metacognition, and applied learning strategies, is critical for student success in these environments.

This study identifies SDL profiles among more than 1,500 undergraduates enrolled in online STEM courses across four broad-access institutions and examines the impact of technology-based interventions – prompts, videos, and peer interaction activities – on students’ likelihood of belonging to each profile over time.

Using model-based clustering, Collaborative researchers identified three distinct SDL profiles: (1) high self-regulation, (2) low confidence but high vigilance around studying, and (3) lower self-reported SDL skills. They found that demographic and academic characteristics, such as race/ethnicity, enrollment status, and Pell eligibility, significantly predicted membership in each cluster, and that students who engaged with the strategies showed greater membership in the high self-regulation cluster over time.

These findings highlight the value of a person-centered approach to SDL and suggest that technology-integrated supports can promote adaptive learning behaviors in online STEM learning environments.

Read the paper: Self-Directed Learning Profiles and the Influence of Technology-Based Interventions Among STEM Undergraduates

 

Categories: Publications Self-directed Learning