affected by individual differences in electrodermal lability in these examinees.

A subsequent DoDPI-sponsored study using a comparison question test (Ingram, 1994) found no relationship between electrodermal lability and the detection of deception by blind scorers. This study also found, however, that the proportion of the subject sample accurately detected as deceptive using skin conductance amplitudes was not significantly above chance. These are the only reports of such associations we were able to find, other than two doctoral dissertations that had other methodological problems and were never published.

We have found no studies of how any other individual differences in psychophysiological responsiveness may affect the accuracy of polygraph tests. In sum, investigation of whether individual differences in physiological responsiveness is associated with the accuracy of polygraph detection of deception has barely begun.5

Individual Differences in Personality

A small body of research addresses the question of whether the accuracy of polygraph testing is affected by the personality traits and characteristics of examinees. The research has addressed some personality traits characteristic of psychologically “normal” individuals and some characteristics of psychologically “abnormal” individuals. Various theoretical rationales have been offered for expecting that the investigated traits might affect physiological responses during polygraph testing.

Studies have been conducted comparing individuals in normal populations who are “high” and “low” on personality dimensions, such as trait anxiety (Giesen and Rollison, 1980), Machiavellianism (Bradley and Klohn, 1987), and self-monitoring (Bradley and Rettinger, 1992). Studies on abnormal individuals have been confined primarily to personality disorders (Gudjonsson, 1982) and psychopathy (e.g., Hammond, 1980; Patrick and Iacono, 1989; Raskin and Hare, 1978). These studies vary substantially in their internal and external validity. All of them were based on specific-incident scenarios, not screening scenarios.

Two studies found that “normal” personality traits moderated physiological indexes of deception. Giesen and Rollison (1980) found that the self-reported trait of anxiety affected skin conductance responsivity during a concealed information test such that subjects with high trait anxiety who were “guilty” of a mock crime responded more strongly than subjects low on trait anxiety. Subjects with low anxiety showed little skin conductance responsivity, regardless of whether they were innocent or guilty. Bradley and Klohn (1987:747) found that subjects high in Machiavellianism (i.e., those “able to focus more directly on the relevant



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement