A Fairy Tale or a True Story? Lie Detection and Leakage

Abstract

Liars may try to act like they are telling the truth, yet these attempts may fail. Ekman (1992) has argued that liars may leak clues to their true feelings in fleeting expressions. This thesis tested this idea. Facial expression differences between university students’ lies and truths were examined with facial recognition software and behaviour rating scales.

Study 1 examined if participants’ (N=102) facial expressions varied as a function of the content that they had viewed, when they honestly or deceptively described a film. When brief episodes of increased muscle tension were isolated in the computer analysis a clear trend emerged for there to be more muscle movement in lies than truths across the face (i.e. brow, eyes, nose and chin). However, the film content (which was designed to elicit feelings of amusement, disgust or no emotional response) did not contribute to this difference. The trend identified in Study 1, was replicated in Study 2, in which participants (n=41) were asked to discuss their opinions truthfully and deceptively. However, the findings did not generalise to a condition where participants (n=40) provided arguments that were consistent or inconsistent with their opinion, when lying was not required. Thus, factors specific to situations in which lying is required (e.g. attempts to manage deception cues) may account for the difference.

Study 3 examined if students with better emotion recognition skills (n=68) may be better at detecting lies because they notice emotional expressions that others may miss. Participants (N=138) rated the behaviour of speakers from Study 1, and judged whether they were lying. The results suggest that lie judgements may be related to the participants’ expectations, and that those with better emotion recognition skills may not have special skills at distinguishing genuine and deceptive expressions. In Study 4, ratings were obtained for messages collected in Study 2 (N=100). Arguments (i.e. opinion-consistent/inconsistent) varied for emotional expression, but lies and truths did not. Thus, it appears impression management processes may contribute to the difference in facial lability identified in Study 1 and 2.