Retrieval Practice in Higher Education: Causality and Content Transfer Effects in a Gateway Math Course

Retrieval practice is a well-established strategy for enhancing memory retention. However, its effectiveness in authentic educational settings—particularly in mathematics and with respect to content transfer—remains understudied. We estimated the causal impact of retrieval practice on student performance in higher education, focusing on task-specific learning gains and content transfer (near vs. far). Participants were three cohorts of first-year business administration students in a gateway mathematics course (n1 = 636 in 2022/2023; n2 = 584 in 2023/2024; n3 = 571 in 2024/2025). Students completed weekly online retrieval practice exercises, with randomized assignments to different problem sets. Performance was assessed through three quizzes and a final exam. We estimated the local average treatment effect (LATE) using an instrumental variable approach combined with individual fixed effects. Retrieval practice improved quiz performance by 4–5 percentage points. Effects were strongest for isomorphic problems, then near-transfer problems, and only limited for far-transfer problems. On the final exam, retrieval practice yielded an 8-percentage-point LATE in one cohort, with smaller, nonsignificant gains in the others. The temporal distance between retrieval practice and quizzes did not systematically moderate outcomes. Our findings provide causal evidence that retrieval practice enhances mathematical learning in higher education by promoting a deeper conceptual understanding and facilitating near, and, in some cases, far content transfer.