Synapse Lab Seminar - Identifying the STEM Learning Technology

The number of annual graduates with STEM degrees has grown dramatically over the past decade. Despite this growth, many students enrolled in STEM struggle to complete their program. An inherent feature of STEM education is that it has a natural cumulative learning structure which makes learning advanced skills in STEM quite challenging. This paper is the first to credibly estimate the cumulative learning technology in a foundational STEM course. Doing so is incredibly challenging as precise measures of effort inputs are typically unavailable and are also dynamic endogenous choice variables. To overcome such challenges, I first gather rich data covering 3,700 undergraduates at a large public university taking an online introductory programming course that has a cumulative structure. The online learning environment serves to monitor students' effort allocation and knowledge accumulation at each stage of the learning process. Then I carry out a field experiment that generates period-by-period exogenous variation in effort allocation, enabling me to identify dynamic interactions across effort inputs in the learning technology. I find evidence of dynamic learning complementarities as the marginal benefit to studying in each learning period is increasing in prior knowledge accumulated. Additionally, I find that effort investment in earlier learning periods results in larger productivity gains for future effort exertion. These results have implications for effective learning strategies and approaches to course design which are further investigated in related work.