NCJ Number
196936
Journal
Polygraph Volume: 31 Issue: 2 Dated: 2002 Pages: 96-107
Editor(s)
Deedra Senter
Date Published
2002
Length
12 pages
Annotation
This article discusses and analyzes the major empirical evidence pertaining to the claims made about voice stress analysis, specifically the assertion that voice stress devices are effective in lie detection.
Abstract
Within the past decade, the voice stress analyzer has received the most notoriety in the field of lie detection and marketed for law enforcement and forensic science purposes. Today, there are some four or five different voice stress analyzers on the market. Voice stress devices extract from the vocal spectrum a subaudible microtremor signal that is seen as useful in detecting stress in a speaker’s voice. Past research on voice stress analysis has claimed the success of its utility and reliability as lie detectors over that of the polygraph. This article reviewed the empirical evidence and the claims made on voice stress devices. It was found that the promise of voice stress analysis in the lie detection field was not and may never be a reality. The evidence showed that none of the devices were useful in detecting deception. The reliable evidence that did exist showed that there was no relationship between what the voice stress devices detected and deception-induced stress. Before analysis of the voice can prove its value in lie detection, it must overcome the significant obstacle of adequately specifying the exact relationship between the components of the voice spectrum and emotional states.