Understanding and Enhancing Self Regulated Learning in Introductory Computer Science Courses

Eric Fouh
Assistant Practice Professor
University of Pennsylvania

Students struggle in introductory programming courses for many reasons, including a lack of prior familiarity with key computer science concepts and skills. Considerable research and development efforts have reduced failure rates in introductory CS courses in recent years, going from 33% to 28% in a ten years period for CS1 courses. However, many students in introductory CS courses struggle because of a lack of self-regulated learning (SRL) skills. There has been work to study learner strategies in introductory computer science through studying programming activities. There has also been work on developing a conceptual understanding of computer science through work with interactive materials. Effectively learning the concepts and skills of introductory computer science calls upon not just one of these two activities but on their intersection. Unfortunately, existing research looks at these activities in isolation. This IUSE project proposed studying and scaffolding the self-regulated learning that emerges at the intersection of these two types of activities. The proposed work aims to address the following driving research questions:● RQ1: What are the usage patterns when students work in online textbooks, programming IDEs, and other learning resources? How do students move between and integrate across these resources?● RQ2: How do students’ learning patterns, identified in RQ1, interact with their self-regulated learning and self-efficacy/confidence development?● RQ3: How do learning patterns, self-regulated learning, and self-efficacy/confidence shape long-term interest and success in CS?● RQ4: How can nudge interventions and reports support the development of SRL and domain learning in CS?When exploring patterns of usage of learning resources before the programming phase (RQ1), we found that most students in a CS1 course spent time (re)watchingvideo recordings of course materials and visiting the course’s Q&A online platform before starting to code. On the other end, fewer students attended office hours or went over the online textbook. We found nonsignificant moderate to strong correlations between precoding activities and the assignment’s difficulty. We also found that the more days spent on the Q&A platform, the earlier they started coding. However, the more posts they viewed (Q&A) and the more minutes they watched (course videos), the later they started coding. And the number of days spent on the Q&A and video recording platforms before coding was positively associated with grades on the homeworks.For RQ2, we found that most students engaged in voluntary practice. However, students with higher prior performance and non-procrastinators were more likely to participate. And participation in the voluntary practice did not significantly impact grades. In addition, we found a weak negative correlation between procrastination and time spent on homeworks. And a weak negative correlation between procrastination and distributed practice. Finally, we found that non-procrastinators performed significantly better than procrastinators on most homeworks.For RQ4, we found that students who received an email nudge (treatment) did not change their help-seeking behaviors. In addition, there was no difference in grades between the treatment and control groups. However, students in the treatment group used more free late days only on the post-nudge homework.


Ryan Baker, University of Pennsylvania, Philadelphia, PA; Wellington Lee, University of Pennsylvania, Philadelphia, PA; Jiayi Zhang, University of Pennsylvania, Philadelphia, PA; Taylor Cunningham, University of Pennsylvania, Philadelphia, PA;Rashmi Iyer, University of Pennsylvania, Philadelphia, PA;