Samenvatting

A single observation of deviant behavior by security officers often provides insufficient information about potential malicious intentions of individuals. This article shows that combining multiple security officers’ observations of deviant behaviors leads to better judgments about possible malicious intentions than judgments of individual officers. In the study described in this paper, experienced supervisors viewed CCTV images of a mall in which pickpockets were active (in reality these pickpockets were confederates). Participants could tag individuals in the footage if they felt that these individuals portrayed deviant behaviors. Using statistics based on signal detection theory, we calculated the accuracy with which individuals in the footage with hostile intentions (i.e., our confederates) were distinguished from individuals without hostile intentions (i.e., the general public). We compared the aggregated individual accuracy scores of participants to accuracy scores of combinations of 2, 3, or 4 participants. Results show that combining observations from two participants leads to higher accuracy than observations of individual supervisors. This effect is mainly caused by an increase in hits (correctly tagged pickpockets). As observations of more participants are added (viz., combinations of 3 or 4 participants) accuracy stagnates. This is mainly caused by an increase in false alarms (judged hostile by participants, but in reality innocent visitors). Implications are discussed for organizations in the security domain and for further research.