A survey: reliability of polygraph examinations conducted by Virginia polygraph examiners

Background & Context

In 1981, Virginia was at the forefront of polygraph regulation in the United States, having implemented one of the nation's first comprehensive licensing laws for polygraph examiners. This licensing framework required high levels of qualifications and training for examiners, creating a unique opportunity to assess polygraph accuracy under regulated professional conditions. Robert H. Edwards, then Assistant Director of the Virginia Bureau of Forensic Sciences and a Backster-trained examiner with 14 years of experience, undertook this survey to determine whether the examinations conducted by Virginia's licensed examiners met scientific standards of reliability.

The study emerged during a critical period in polygraph history. Research conducted since the 1923 Frye decision showed polygraph accuracy in the mid-90 percent range, yet courts remained skeptical about admitting polygraph evidence. This study was prepared for the Commonwealth of Virginia and distributed to all Congressmen, reflecting its intended role in shaping policy debates about polygraph admissibility and professional standards. Virginia's licensing system provided a natural laboratory for evaluating whether strict credentialing requirements translated into improved field performance.

Research Design & Methodology

A questionnaire was mailed to each licensed polygraph examiner residing in Virginia, targeting practitioners across both law enforcement and private sectors. Information was requested for specific examinations conducted during 1980, including the categorizing of examination results by the determinations rendered, and the number and category of verified determinations. This retrospective survey design focused on real-world field examinations rather than laboratory simulations, examining cases where ground truth could be established through independent verification.

A total of 147 questionnaires were mailed, yielding a response rate of 48.3 percent overall. However, participation varied significantly by examiner type: a total of 77.5 percent of the law enforcement examiners responded, indicating substantially higher cooperation from the public sector. The study employed a descriptive survey methodology, relying on examiners' self-reported data about their examination outcomes and subsequent verification of those determinations through confessions, independent evidence, or other corroborative means. The study addressed the reliability of "specific" examinations of a criminal nature.

Key design elements included:

• Survey of all licensed Virginia polygraph examiners (N=147)
• Focus on criminal examinations conducted in 1980
• Self-reported accuracy data with external verification
• Separation of law enforcement and private examiner results

Results & Key Findings

The headline finding was remarkably optimistic: the mean accuracy level for Virginia law enforcement examiners was found to be 98.3 percent. This extraordinary accuracy claim was based on a substantial dataset from criminal investigations.

There were 16 errors in the 959 verified law enforcement examinations, with the mean error being 1.7 percent. This represented nearly 1,000 real-world criminal cases where polygraph determinations could be checked against independent evidence. The verification process relied on confessions, physical evidence, witness statements, and other investigative outcomes that established ground truth.

Additional findings:

• 77.5% response rate from law enforcement examiners
• 48.3% overall response rate across all examiner types
• 16 total errors out of 959 verified examinations
• 1.7% error rate for law enforcement examiners

The study's findings aligned with the broader literature review Edwards conducted, which showed the polygraph to have an accuracy level in the mid-90 percent range across multiple studies since 1923.

Discussion & Significance

Edwards' findings represented some of the highest accuracy rates reported in field validity research up to that time. The 98.3% accuracy figure substantially exceeded the mid-90% range typical of prior research, suggesting that Virginia's licensing and training requirements may have elevated examiner performance. The study was cited in the landmark 1983 Office of Technology Assessment report on polygraph validity and became part of the evidence base that polygraph proponents used to argue for legal admissibility and expanded use.

Based upon the findings of this study, Edwards recommended that the courts accept polygraph results as circumstantial evidence after determining examiner qualifications and the technique used. This recommendation emphasized the importance of examiner credentials—a theme central to Virginia's regulatory approach. The study suggested that with proper training, licensing, and professional standards, polygraph testing could achieve reliability levels approaching those of other forensic sciences.

The research had significant policy implications. Virginia's licensing model, which the study recommended continuing, became a template for other states considering polygraph regulation. The high accuracy rates provided ammunition for polygraph industry advocates seeking to expand the use of testing in criminal justice, employment screening, and security vetting contexts during the 1980s.

Limitations & Considerations

The study's methodology presents several important limitations. The reliance on self-reported data from examiners creates potential for selection bias—examiners may have been more likely to report and verify cases where they were confident in their determinations. The 48.3% overall response rate raises questions about non-response bias; examiners with poorer accuracy may have been less likely to participate. Additionally, the verification methods were not standardized or independently validated, and the study did not specify how "errors" were definitively established or whether inconclusive results were included in accuracy calculations.

Critical methodological concerns include:

• No independent verification of examiner-reported outcomes
• Possible confirmation bias in case verification processes
• Unclear treatment of inconclusive results
• Limited information about examination techniques used
• No control for case difficulty or examiner experience levels

The study's optimistic findings contrasted sharply with later critical assessments. The 2003 National Academy of Sciences review of polygraph research would identify significant methodological weaknesses in many field validity studies, including problems with verification criteria and outcome reporting. Studies relying on examiner self-reports and case verification through confession have been particularly criticized for potential circularity—confessions may be influenced by polygraph results themselves.

Practical Applications

For polygraph practitioners, the Edwards study underscored the potential value of formal training, licensing requirements, and adherence to standardized procedures. The superior response rate and performance of law enforcement examiners suggested that institutional support, ongoing quality assurance, and professional accountability structures may enhance examiner reliability. The study provided empirical support for states considering polygraph licensing laws and for agencies establishing internal quality control programs.

For consumers and policymakers, the research highlights the importance of examiner credentials and regulatory oversight. While the study's accuracy claims should be interpreted cautiously given methodological limitations, it demonstrated that systematic tracking of verification outcomes is both feasible and valuable for assessing examiner performance. The Virginia model—combining licensing, training standards, and professional accountability—offered a framework for quality assurance that extended beyond the specific accuracy figures reported.