NCJ Number
89713
Journal
Journal of Police Science and Administration Volume: 11 Issue: 1 Dated: (March 1983) Pages: 62-68
Date Published
1983
Length
7 pages
Annotation
This test of the effectiveness of the Psychological Stress Evaluator (PSE), which claims to detect signs of emotional stress in the human voice, in detecting deception shows the PSE to be unreliable.
Abstract
The PSE analyzes audio tape recordings and, according to its manufacturer, detects inaudible frequency modulations in the 8-12 Hz region caused in part by micro-tremors of the muscles associated with the voice production area. A reduction of those tremors is said to result in greater stability of the vocal frequencies being emitted, which the manufacturer claims are both indicative of stress and accurately measured by the instrument. In this test, the subjects were six undergraduate students (three males and three females) who were facing university disciplinary action after having had their cases reviewed by the same university disciplinary dean. The students were assigned to one of two groups based on their infractions. Three students accused of smoking marijuana, who neither confessed nor refused to take the test, composed one group. The second group faced disciplinary action for other offenses, but were asked questions pertaining only to an incident of marijuana smoking in a particular classroom. Audio tapes were sent to a PSE analyst. Based on the charts for the six subjects, the analyst made one correct, three incorrect, and two inconclusive judgments. Thus, the PSE failed to perform at a level better than chance expectancy. The poor showing of the PSE in this study appears to be a reflection of the instrument itself rather than in the procedures used to evaluate it. Tabular data and 18 references are provided.