Decomposing a pre- post-test outcome to measure the effect of cooperative learning on student achievement

Abstract

Using a quasi-experimental research design, we examine the efficacy of one type of cooperative learning pedagogy (think-pair-share exercises) in the microeconomic principles course. Students in the treatment group engaged in think-pair-share exercises with an assigned partner throughout the semester. Those in the control completed the same exercises but did so individually. All students took the Test of Understanding in College Economics (TUCE) at the beginning and end of the semester to allow us to determine any achievement gains and whether these gains differ by treatment. Following Walstad and Wagner (2016) and Smith and Wagner (2018), we decompose student performance to consider positive learning – adjusting for guessing – on the TUCE. We find no significant effect of cooperative learning on guessing-adjusted positive learning measures.