Statistical versus clinical lie detection reconsidered.

Responds to criticism by G. Ben-Shakhar and I. Lieblich (see record 1984-24962-001) regarding the present authors' (see record 1982-01946-001) study on statistical and objective analysis of polygraph charts. It is agreed that although the test is not likely to be valid for most of the situations in which it is used, the statistical formula may beat the clinician in interpreting some physiological protocol data in some circumstances. (7 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Lie detection in ancient and modern times: A call for contemporary scientific study.

Notes that psychologists have neglected investigating polygraphic lie detection, a technique that is seen as clearly within their purview. It is basically a psychological test, although with questionable psychometric merit, that assumes that liars are aware of their lying, which in turn causes measurable emotional reactions. This simplistic assumption was not always shared by the ancients, but it now has widespread contemporary acceptance. The polygraphic technique based on this assumption yields unacceptably high error rates that have had ruinous effects on the lives of many misclassified truthful persons. It is argued that a recent executive order to use lie detection to ensure national security, if left unchallenged by informed psychological opinion based on sound scientific research, opens the door even wider for further misuses of a psychological tool. (60 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Statistical versus clinical lie detection.

Argues that studies conducted to date have tended to ignore the single most important variable of polygraph validity: the human judge. The present study focused on clinicians' interpretations of polygraph protocols and showed that clinicians performed less accurately than statistical analyses. Ss included 30 undergraduates administered a polygraph test by 4 examiner-trainees and 30 experienced polygraph interpreters. Statistics outperformed human judges because they used information optimally and applied decision rules consistently, while clinicians tended to add error variance to their protocol interpretations. Unfortunately, current empirical evidence suggests that the prospects for improving clinicians' consistencies are not promising; the possibility of applying statistical methods to interpreting polygraph data is recommended. It is suggested that psychologists become more active researchers in this area, a domain that is properly within their scientific purview. (41 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)