NCJ Number
105379
Journal
Jurimetrics Journal Volume: 27 Issue: 3 Dated: (Spring 1987) Pages: 263-270
Date Published
1987
Length
8 pages
Annotation
This article examines the validity (accuracy) of urinalysis to detect the consumption of illegal drugs and the validity of the polygraph in testing for deception.
Abstract
The validity of a test refers to the probability that the test result is accurate or true. In the case of urinalysis, when screening large numbers of persons of whom only a small proportion are offenders, the validity of the subset of positive results will be much lower than the advertised validity of the screening test. The majority of persons who fail the screening test can easily be innocent victims of false-positive test errors. In the case of the polygraph, when a large number of persons are tested, of whom the majority are offenders, the validity of negative or exculpatory polygraph tests will be much lower than the overall validity of the technique. If the polygraph test is administered adversarily, if the suspects are naive (do not know the technique for beating the polygraph), and if about half of all the persons tested are liars, then polygraphers will be wrong about one-third of the time overall and about one-half the time when testing innocent suspects. Should any of the 'if' conditions not be met, the polygrapher's validity will be worse, a result worse than random guessing. These findings preclude the use of a polygrapher as an expert witness for either the prosecution or the defense. 1 table and 13 footnotes.