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

Classroom experiments are a popular tool among economics instructors. A rich experimental literature studies their impact on student learning with between-subject designs that randomize classroom experiment use across course sections. Such designs are difficult to set up and conduct in a way that achieves adequate statistical power. We complement this literature with a within-subject experiment where classroom experiments are randomized across topics within a section, which enhances power and is easier to implement. We find little impact of our classroom experiments on student achievement for either male or female students, but these null results may mask heterogeneous effects across topics.