Lyin' eyes: Ocular-motor measures of reading reveal deception.

Our goal was to evaluate an alternative to current methods for detecting deception in security screening contexts. We evaluated a new cognitive-based test of deception that measured participants' ocular-motor responses (pupil responses and reading behaviors) while they read and responded to statements on a computerized questionnaire. In Experiment 1, participants from a university community were randomly assigned to either a "guilty" group that committed one of two mock crimes or an "innocent" group that only learned about the crime. Participants then reported for testing, where they completed the computer-administered questionnaire that addressed their possible involvement in the crimes. Experiment 2 also manipulated participants' incentive to pass the test and difficulty of statements on the test. In both experiments, guilty participants had increased pupil responses to statements answered deceptively; however, they spent less time fixating on, reading, and rereading those statements than statements answered truthfully. These ocular-motor measures were optimally weighted in a discrimination function that correctly classified 85% of participants as either guilty or innocent. Findings from Experiment 2 indicated that group discrimination was improved with greater incentives to pass the test and the use of statements with simple syntax. The present findings suggest that two cognitive processes are involved in deception-vigilance and strategy-and that these processes are reflected in different ocular-motor measures. The ocular-motor test reported here represents a new approach to detecting deception that may fill an important need in security screening contexts.