Predicting Known Terrorist Event Outcomes and the Illusion of Self‐Assessed Expertise*

Abstract

Objective This study focuses on the influence of self‐assessed expertise on forecasting accuracy for incomplete information. We investigate this ability in individuals using a series of indicators of terrorist behaviors. Method An instrument containing incomplete data on terrorist attacks was provided to participants. Participants were instructed to predict missing values. The accuracy of approximately 1,900 forecasts was determined. The influence of self‐identified statistical ability, terrorism familiarity, and an empirical measure of pattern recognition ability on accuracy was investigated. Results The results of the experiment indicate that neither self‐assessed statistical ability nor terrorism familiarity have a significant influence on forecast accuracy. Interestingly, the empirical estimate of pattern recognition ability also had no appreciable effect on accuracy. Conclusion This study indicates that statistical ability and terrorism familiarity have little effect on forecast accuracy. Innate pattern recognition ability also showed no significant relationship with accuracy. These findings suggest that self‐identified expertise should initially be viewed with some level of skepticism.