The minutes of the 18 July 2023 meeting of the state of Virginia’s Polygraph Examiners Advisory Board reveal that “in March of 2023 Board staff attended a demonstration for Computer Voice Stress Analyzer (CVSA). CVSA software programs are designed to measure changes in voice patterns caused by the stress, or the physical effort, of trying to hide deceptive responses. CVSA programls interpret changes in vocal patterns and indicate on a graph whether the subject is being ‘deceptive’ or ‘truthful.'”
Under Virginia law, anyone offering lie detection services must be licensed by the state, and the use of any lie detection device other than the polygraph must be approved by the Director of the Department of Professional and Occupational Regulation.
The minutes of the Polygraph Examiners Advisory Board show that its members unanimously voted to reject the use of CVSA:
After discussion, a motion was made by Mr. Holzbach and seconded by Mr. Grant based on a review of the public comments, the presentation by the CVSA manufacturer, NITV Federal Services, a review of the current scientific evidence, and independent research by individual board members, the Board recommends that the Director not approve CVSA as a deception detection or truth detection device. The preponderance of the current scientific literature indicates that CVSA does not collect reliable and valid data on detecting deception and verifying truthfulness, which was confirmed during NITV’s inconsistent demonstration of the product. The Board is willing to accept new scientific evidence of CVSA’s reliability.
The motion was approved with a vote of 5-0-0 Ayes: Lovely, Orr, Brown, Holzbach and Grant. Nays: None Abstained: None Absent: None
It is not surprising that a public advisory board dominated by polygraph operators (Chairman Michael E. Lovely and members Heather Brown, Karl H. Holzbach, and John T. Orr are all polygraphers) would reject the competing pseudoscience of voice stress analysis. However, were they to assess polygraphy with the same rigor they apply to voice stress analysis, they would have to likewise conclude that the preponderance of the current scientific literature indicates that the polygraph “does not collect reliable and valid data on detecting deception and verifying truthfulness.”