Finding a needle in a haystack: Toward a psychologically informed method for aviation security screening

Current aviation security systems identify behavioral indicators of deception to assess risks to flights, but they lack a strong psychological basis or empirical validation. We present a new method that tests the veracity of passenger accounts. In an in vivo double-blind randomized-control trial conducted in international airports, security agents detected 66% of deceptive passengers using the veracity test method compared with less than 5% using behavioral indicator recognition. As well as revealing advantages of veracity testing over behavioral indicator identification, the study provides the highest levels to date of deception detection in a realistic setting where the known base rate of deceptive individuals is low. Identifying threats presents a huge challenge to those tasked with ensuring public safety, and to psychologists developing methods for detecting deception. However, the news from both arenas is not good. Since the events of September 11, 2001, billions of dollars have been invested in aviation security procedures designed to detect threats to airplanes (United States Government Accountability Office, 2011), but the effectiveness of these procedures has been questioned Threat detection procedures typically involve looking for individuals who display behaviors thought to be indicators of deception, particularly behaviors shown by the perpetrators of previous attacks. The effectiveness of behavioral indicator approaches has never been tested in a large-scale field trial. However, a metaanalysis of laboratory studies that used behavioral indicators to discriminate deceivers from truth-tellers revealed a mean rate for correct identification of only 54% In this article, we present a new procedure for aviation security screening that is based, not on behavioral indicators selected from previous incidents, but on testing the veracity of passengers' verbal accounts. Our approach takes techniques derived from psychological theory and shown in recent laboratory studies to yield promising rates of deception detection, and integrates them into a comprehensive procedure for detecting threat. We then compare the effectiveness of behavioral-indicator and veracity-testing approaches in an in vivo empirical evaluation conducted with passengers departing on flights at international airports during routine security screening. Approaches to Detecting Deception The majority of published research on detecting deception has set out to identify indicators in human behavior that can discriminate deceivers from truth-tellers. Behavioral indicators of deception fall into two main categories: physical behaviors relating to demeanor (e.g., nervousness, aggression) and/or actions (e.g., eye contact, fidgeting); and verbal behaviors relating to the nature and production of speech (e.g., hesitations, use of pronouns). As noted above, behavioral indicator approaches typically yield low discrimination rates. Indeed