Eye tracking & detection of deception: A pupillometry approach to the Concealed Information Test

Abstract

The Concealed Information Test (CIT) is one of the most scientifically supported techniques of lie-detection to date, and it is specifically used for detecting individuals' knowledge of crime details. Advancements in the field of eye-tracking research also indicate that pupillometry data could be of value for detection of deception techniques. Recognising a research gap regarding the potential benefits of using eye-tracking data for detection of concealed information in CIT studies, the present study investigated whether eye fixation duration (i.e. dwell time) and average pupil size (APS) could predict individuals' attempts to conceal information in a CIT procedure that used mock-crime photographs as its primary source of test stimuli. A total of 30 participants took part in a mock-crime CIT experiment and were assigned to two groups (guilty and innocent). Guilty participants were required to enact a mock-theft, whereas innocent participants performed an observation task. The CIT had two types of question trials (critical and control-relevant trials). Critical trials presented the simultaneous display of crime-relevant and crime-irrelevant photographs, and correct answers could only be known to guilty participants. In control-relevant trials, the correct answers could only be known to innocent participants, and they showed the simultaneous display of control-relevant and control-irrelevant photographs. Results revealed that guilty participants displayed significant increases in APS when responding to both critical and control-relevant trials, in comparison to innocent participants. Moreover, neither guilty nor innocent participants displayed significant differences in APS between responses towards relevant and irrelevant photographs. Regarding dwell time responses, results showed that only guilty participants presented significant decreases in mean dwell time towards relevant(i.e. known)photographs, compared with irrelevant photographs in critical trials. By contrast, innocent participants displayed significant increase in dwell time towards relevant photographs in comparison to irrelevant photographs in control-relevant trials, whereas guilty participants presented significant decrease in dwell time towards relevant photos in these trials. These results suggest that guilty participants engaged in avoidance behaviour during the CIT examination, and detection of participants' attempts to conceal information is possible by targeting guilty participants' eye-movement patterns. Moreover, data from innocent participants suggest that innocent participants are less likely to engage in avoidance behaviour during CIT examinations and they may display an orienting response towards known test stimuli instead.