Detection of Deception with Event-Related Brain Potentials

This article reviewed previous studies concerning psychophysiological detection of deception in which event-related brain potential(ERP) had been recorded as a response index. The usual dependent measures in detection studies are autonomic activities such as respiratory, electrodermal, and cardiovascular responses that are assumed to be correlates of the emotional state induced by crime relevant questions. Recently, however, researchers in this field have emphasized the importance of cognitive factors rather than emotional ones to detect guilt and this trend facilitates the use of ERP as a detection measure.   Several ERP components have been identified and each functional significance during human information-processing has been well described. In the ERP-based detection studies, a component such as P3, N400, or CNV(contingent negative variation) have been used as a detection index. The P3 is elicited by rare and meaningful events that are relevant to the participant's task. The N400 component is elicited by semantic anomalies such as words that complete sentences falsely. The CNV has been variously described as related to expectancy, motivation, and attention. Although these three ERP components seem valid detection measures, the most promising index in the field practice is the P3 component because of its relative ease of recordings and large amplitudes. Moreover, in laboratory studies, correct detection rates by the P3 component are slightly higher than the rates by the conventional autonomic measurements.   Several merits of using an ERP component as a detection index are summarized as follows: 1) advancement and elaboration in judgment of test results because ERP are specifically time locked to an event, 2) an improvement of the correct detection rate, 3) an improvement of test objectivity and reliability, and 4) counteracting countermeasures. Before the practical use of ERPs in the psychophysiological detection of deception, however, a method of artifact-free measurement of ERPs and a definite criterion of judgment for individuals must be established.