Deception detection process and accuracy: an examination of how U.S. military officers detect deception in the workplace

Abstract

Research shows that humans are, on average, only slightly better-than-chance at deception detection. Meta-analysis conducted by Charles Bond and Bella DePaulo in their work Accuracy of Deception Judgments published by Personality and Social Psychology Review in 2006 yields an across-study average accuracy rate of 54%. Although prior research has failed to identify variables that have a large impact on accuracy, a recent set of studies focused on diagnostic utility (strategic questioning) leads us to expect substantial question effects producing levels of accuracy that differ substantially from chance. Recent research advocated for abandoning cue-based deception detection in favor of the idea of diagnostic utility. Specifically, this new line of research provides a basis for specifying the conditions under which questioning of honest and deceptive individuals yields levels of deception detection accuracy that depart substantially in both directions from the usual slightly-better-than-chance results that characterize past attempts. This thesis is a replication of these most recent diagnostic utility studies to determine if the methods are (1) generalizable to a new population and (2) useful in identifying specific questioning strategies relevant to Department of Defense and fraud investigation activities.