VOLUME 5 ISSUE 3/4 FALL 2007 1 THE CANADIAN JOURNAL OF POLICE & SECURITY SERVICES VOLUME 5 ISSUE 3/4 FALL 2007 Concealment of Guilty Knowledge To the best of my knowledge, Officer: Investigation of Crime Questionnaire Patterns Associated with Guilty Knowledge Concealment Stephen Porter* Dalhousie University Catherine Stewart University of Saskatchewan Mary Ann Campbell, Ph.D. University of New Brunswick-Saint John A ccuracy and Honesty: Investigation of Crime Questionnaire Patterns Associated with Guilty Knowledge Concealment The statements elicited from suspects during police interviews are essential pieces of evidence in solving crimes. Investigators must decide whether and when a particular suspect is lying or telling the truth, a task known as credibility assessment. Of course, many of the stories offered by (guilty) suspects in the context of an interrogation are lies, such as false denials and alibis. With this knowledge, most police interviewers receive extensive training in deception detection and are highly motivated to use their training to catch liars. Sometimes, however, traditional police training may not enhance, and may even impair, the ability to accurately assess credibility (e.g., Kassin & Fong, 1999; Meissner & Kas- sin, 2002; Vrij, Mann, & Fisher, 2006). For example, some police training encourages the use of emotion-based, stereotypical signs of lying such as “shifty eyes”, gaze aversion, and nervous gestures, rather than empirically based cues to deception (e.g., Mann, Vrij, & Bull, 2004; Strömwall & Granhag, 2003; Vrij, 2004). Even empiri- cally-based training programs aimed at improving de- ception detection ability have produced only modest im- provements in performance (see Masip, Sporer, Garrido, & Herero, 2005; Vrij, Evans, Akehurst, & Mann, 2004). In examining the scientific literature, the most success- ful deception detection training program to date was by Porter, Woodworth, and Birt (2000) who provided evi- dence-based training to a group of Canadian parole of- ficers over a two-day workshop. By the end of training, the ability of the parole officers to detect deception in videotaped speakers had improved significantly, from an average of 40.4% accuracy at baseline (performance that, alarmingly, was below chance) to 76.7% by the end of training. However, in general, police, judges, and other legal decision-makers generally perform around the level of chance in detecting deception (e.g., Bond & DePaulo, 2006; Ekman & O’Sullivan, 1991; Granhag & Stömwall, 2005). Further, it is possible that investigators *Correspondence concerning this article should be addressed to: Ste- phen Porter, Ph.D., Psychology Department, Dalhousie University, Halifax, Nova Scotia, Canada, B3H 4J1 (email: sbporter@dal.ca). This study was supported by grants to the first author from the Social Sciences and Humanities Research Council of Canada (SSHRC) and the Natural Sciences and Engineering Research Council of Canada (NSERC). We thank Andrea Beck, Alaya Khatkhate, Janet Pyke and Valerie Young for volunteering in the present study. ABSTRACT Determining the credibility of information obtained from suspects during police interviews is an essential aspect of crime-solving. This study examined the cognitive strategies used by subjects who attempt to conceal their knowledge of a crime, and whether deception in this context is detect- able. Undergraduate participants were randomly assigned to one of three groups: (1) perpetrators who committed a mock theft; (2) scene witnesses who had knowledge of crime scene characteristics but not the crime; or (3) naïve respond- ers who had no knowledge of details concerning the crime scene or of the crime itself. All participants then completed a Guilty Knowledge Questionnaire (GKQ), created for this study. Students acting as deceptive perpetrators were in- structed to respond to questionnaire items in any manner that they felt would help them evade detection, while the others were instructed to respond honestly. As expected, deceptive perpetrators responded less accurately on the total GKQ than honest perpetrators and scene witnesses, but did not differ significantly from naive responders. When crime scene and theft-knowledge questions were considered separately, however, deceptive perpetrators (unlike naïve responders) selectively acknowledged some awareness of scene details while minimizing knowledge of the crime itself.