The Effects of a Problem-Solving Class on Student Persistence in STEM

Student persistence in STEM has been a focus of educational research, with both quantitative and qualitative methods being used to investigate patterns and causes of attrition. Some studies have used machine learning to predict a student's likelihood to persist given measurable classroom factors and incoming preparation, while others have framed persistence as a function of a student's social integration in the classroom. While these methods have provided insight into underlying causes of attrition in STEM, they have not investigated class structures or teaching methods that promote persistence. In this study we explore how an introductory active learning physics class using real world problem-solving (PS) had a positive impact on persistence for students at a large research-intensive university. Our findings showed that the one-year persistence rates for the PS course were 74% (fall) and 90% (spring), while the control class had a persistence rate of 64% and 78%, respectively. In spring, the PS persistence rate was significantly higher (p=0.037). The PS also had higher final grades and larger learning gains than the control despite lower incoming preparation. This study motivates future work to understand the mechanisms that promote student persistence in introductory physics.