Classroom Experiments

Using the Experiments

The purpose of the classroom experiments is to show students how some canonical experiments work, and to put students in the position of subjects who need to think through the consequences of different choices.

All the experiments are designed to be run by pen and paper in the classroom. Results are tabulated by entering the results into Excel spreadsheets. This can be done by an assistant while the lecture continues, or ‘live’ with the spreadsheet displayed on a screen. The ‘live’ option usually doesn’t take that long, and students seem to enjoy watching the results (and commenting on strategies/choices) as they come in.

All the available experiments can accommodate at least 48 participants (50, 50, 80 and 48 respectively); you will need to modify the spreadsheets to accommodate more. In larger classes, the experiments seem to work fine if only a random sample of the class is able to participate (which also saves time).

In most cases I pay students with individually-wrapped candies. (Halloween candies work very well if you teach in the Fall!) Other possibilities are small amounts of cash or course credit. Since the experiments are purely illustrative, they can of course be run without any explicit incentives at all.

Complete materials for running four classroom experiments are currently available on Oxford University Press's Ancillary Resource Center (ARC) site. They are (with the published experiment they illustrate):

Experiment 2.1: Gift Exchange (Fehr, Klein and Schmidt, Econometrica 2007).

Experiment 2.2: Costs of Control (Falk and Kosfeld, AER 2006)

Experiment 4.1: Sabotage in Tournaments (Carpenter et al., AER 2010)

Experiment 5.1: Linear Public Goods Game (Fehr and Gächter, Nature 2002)