Using experimental economics to evaluate alternative subjective elicitation procedures

Managing uncertainty is an unavoidable challenge in a variety of decision contexts. If empirical data are available, statistics can be used to assist decision-making. If objective data are absent, experts are commonly used as a source of subjective probability estimates about the variable of interest. In our controlled economic experiment, a web-based software was developed and incentive-compatible rewards were applied to participants in order to induce rational decision-making. Participants assessed data drawn from known distributions and then responded to a subjective probability elicitation. The empirical results provide experimentally based guidance on elicitation and aggregation techniques that yield the most accurate subjective estimates of unknown probabilities. The results can be applied to a wide array of decision-making processes involving subjective probability analysis.