In the blink of an eye: Quantitative blink dynamics predict deceptive personality traits in forensic interviews

Abstract

It is commonly believed that eye movements and blinks can be indicative of deception, however, there are few laboratory studies of such effects in general and no studies in real-world forensic populations in particular. Here we test the hypothesis that automated eye blink detection and blink frequency dynamics would predict trait levels of deception in a forensic sample. Participants were incarcerated male offenders (n = 125) undergoing life history interviews which were used in part to assess psychopathic personality traits using the Hare Psychopathy Checklist-Revised. Psychopathy is a personality disorder characterized in part by trait dishonesty, manipulation, and pathological lying. We developed a novel automated image analyses approach using hidden Markov machines, where the assumption is that an underlying hidden process influences the observable, measurable process (i.e., deception). Consistent with hypotheses, blink dynamics were predictive of interpersonal psychopathic traits (i.e., pathological lying and conning/manipulative behavior). This study provides the first evidence that blink dynamics from a naturalistic forensic interview, may help understand individual differences in trait deception.