Gaze behaviour and deception detection: Insights from an eye tracking analysis of interviewers and observers

This study addresses the potential benefits of eye-tracking to assess the relationship between people’s presuppositions about how liars behave and where people actually look when making deception judgements, and how these presuppositions and looking behaviours are related to judgement accuracy. Participants (N = 68) took on the role of either interviewer or observer in a 2×2 between-subjects design (face-to-face vs. video; active vs. passive role). Gaze data were collected using Tobii Pro Glasses 2 and mapped to predefined facial and bodily areas of interest, e.g. arms and hands, eyes and brows, forehead. Participants’ beliefs about nonverbal deception cues (e.g., gaze aversion, blinking, fidgeting) were collected via questionnaire. Veracity judgments were assessed using both direct ratings and indirect (intuitive) measures. The preliminary findings suggested that looking more often at someone’s legs and feet associated with more accurate deception judgements than when these areas were looked at less often. Looking at the forehead more often leads to lower deception judgement accuracy. Interestingly, while certain presuppositions about deception, such as the belief that liars avoid eye contact, were reflected in participants’ gaze behaviour (e.g., spending more time looking at the eyes), these beliefs and gaze behaviour did not translate into higher judgement accuracy. Looking the deceiver more in their eyes did not predict accuracy. Instead, even when gaze behaviour aligned with such presuppositions, it offered no advantage in detecting deception and often had no predictive value for veracity judgments. Our findings demonstrate a new method for directly testing theoretical associations between people’s beliefs about deception and their actual behaviour when making judgements. Tentatively, our preliminary findings support theories that suggest nonconscious and indirect deception judgements are more accurate than direct deception judgements, due to people attending to unreliable cues to deception, such as eye contact, when making direct judgements.