The Effect of Teaching Economics with Classroom Experimens: Estimates from a Within-Subject Experiment

We build upon the literature that stud- ies the effect of classroom experiments on student achievement in economics courses by employing an within-subejct design that has greater statistical power than previous studies. We find limited evidence of class- room experiments on immediate under- standing across all lessons, but as Durham et al. (2007) also find, the effects vary sub- stantially by lesson. Some classroom exper- iments have large, positive effects on stu- dent achievement, while others appear to have substantial negative effects. Our esti- mates that pool effects across topics mask these differences. We thus conclude that instructors should take care when deciding whether to use a particular classroom ex- periment, as it seems certain lessons lend themselves better to the use of classroom experiments than others.