How Well Do Hypothetical Scenarios Map to Reality? Ann Application of Bayesian Truth Serum

How Well Do Hypothetical Scenarios Map to Reality? Ann Application of Bayesian Truth Serum

A frequent data collection instrument in recent years for criminologists is the hypothetical crime scenario. In this instrument respondents are asked to read a hypothetical crime scenario and respond to questions that include perceptions of the certainty, severity, and swiftness of punishment should they commit the crime described in the scenario and the estimated probability that they would commit the crime described in the scenario condition. There is little research on how accurate these responses are. This project examines this issue by having students anonymously respond to hypothetical scenarios under various experimental conditions: (1) a control condition where no incentive is given (this is the normal condition in previous research), (2) a condition where respondents will be given an incentive to fully complete the instrument, and (3) the Bayesian Truth Serum (BTS) condition. The Bayesian ‘truth serum’ (BTS) is a scoring method that provides truth telling incentives for respondents answering multiple-choice questions about intrinsically private matters (opinions, tastes, past behavior; see Prelec, 2004).

Project Statistics
Principal Investigator
Co-Investigators
Researchers