At our seminar this afternoon Debra Shepherd and Volker Schoër of Wits treated us to some interesting microeconometrics on the economics of education.
First year economics can be a tough course marked by poor pass rates. Tutorial classes are often seen as a cost-effective way to improve students’ performance. Debra argued that the academic literature on this is not conclusive. A simple cross-section regression model where the exam mark is a function of, amongst other things, tutorial attendance does not control for unobserved fixed effects and a randomized control trial or a longitudinal study is required to be able to talk about causality or impact.
They used a fuzzy discontinuity regression design to analyse the results of a quasi-experiment of tutorials and exam performance in the first-year microeconomics class at Stellenbosch University. There tutorial classes were offered to everyone, but they were compulsory for the students that failed the first class test. This left them with two factors that affected exam performance: test scores and the fact that the test performance determined how many tutorials the students attended. There were also two interesting groups to examine: those that just passed the first test and could choose to attend tutorials, and those that just failed it and were forced to attend tutorials. The rest of the analysis gets complicated but suffice to say, it seems that tutorial attendance mattered. Compulsory attendance of every tutorial class added 1.5 percentage points to the exam score for every class attended, compared to those that attended none of the tutorials.
We would like to thank Debra and Volker for making the trip and inspiring some economics education research in our School as well.