Vecon Lab Bayes' Rule Experiment: Introduction

This program sets up an individual decision-making task in which each person obtains a private signal about which of two events has ocurred ("Red" or "Blue") and must report a probability for one of the events. Earnings are structured so that the optimal decision is to report one's true subjective probability, in a DeGroot/Marshack type of mechanism. This setup allows one to evaluate the extent to which reported probabilities match those implied by Bayes' rule. The elicitation procedure is explained more fully in Chapter 30 of Holt (2006) Markets, Games, and Strategic Behavior, Addison-Wesley.

Elicited probability assessments show over-estimation of low probabilities in some treatments. The experiment can be used to illustrate the importance of combining different sources of information and the possible effects of psychological biases.

Vecon Lab - September 5, 2008