Polygraph Validity and Reliability: A Review

Review of laboratory and field studies establishing empirical support for polygraph validity.

Laboratory experiment using mock crime or controlled paradigm with random assignment to guilty and innocent conditions. Physiological responses recorded during CQT/CIT administration.

Published in 1976 in In Ansley (Ed.), Legal Admissibility of the Polygraph, Charles C. Thomas, this work by Barland and colleagues examined polygraph validity and reliability: a review. Key findings include: Major review of laboratory and field studies establishing empirical support for polygraph validity. Identified key factors affecting accuracy and proposed standards for practice. The controlled laboratory conditions ensure known ground truth, enabling precise accuracy measurement.

This foundational review helped establish the framework for scientific evaluation of polygraph that continues to guide the field.

Background & Context

The mid-1970s represented a critical juncture for polygraph science. Despite decades of operational use in law enforcement and security settings, congressional hearings had concluded that federally and privately funded research "had done little to persuade" policymakers of the polygraph's validity. Against this backdrop of skepticism, the National Institute of Justice commissioned Gordon Barland and David Raskin to conduct the first comprehensive scientific evaluation of polygraph validity in actual criminal investigations.

This 1976 work emerged from a multi-year research program funded under contract 75-NI-99-0001. The Barland and Raskin study became one of the ten field studies ultimately included in major scientific reviews, establishing itself as a foundational reference that would shape polygraph policy and research methodology for decades. The study aimed to bridge the gap between laboratory findings and real-world applications, providing empirical data on how polygraph techniques performed with actual criminal suspects whose liberty was at stake.

Research Design & Methodology

The Barland and Raskin field study was among ten major field studies that examined polygraph validity using real criminal cases. Unlike laboratory studies with mock crimes, this research evaluated actual polygraph examinations conducted with criminal suspects in authentic investigative contexts. The researchers employed multiple verification methods to establish ground truth, though cases were considered "verified" primarily by confession of the guilty suspect, with most also corroborated by additional evidence.

A distinctive methodological feature was the use of blind chart evaluation. Polygraph charts were reinterpreted by examiners who were "blind to" the suspect's guilt or innocence, with the degree of agreement serving as a test of validity. This approach attempted to control for examiner bias and interpersonal expectancy effects. The study also compared original field examiner decisions with blind evaluations, though analyses typically used results from original examinations rather than blind evaluations due to concerns about interpersonal expectancy effects reflected in the charts themselves.

• Field examination of criminal suspects in authentic investigative contexts
• Multiple verification criteria including confession and corroborating evidence
• Blind chart re-evaluation by independent examiners
• Comparison of original versus blind evaluation accuracy

Results & Key Findings

The study revealed substantial variability in polygraph accuracy, with correct guilty detections ranging widely and correct innocent detections showing even greater variance, from as low as 12.5 percent in judicial outcome analysis to highs exceeding 94 percent in other conditions. This variability highlighted the complexity of field validation and the critical importance of verification methods and examiner decision criteria.

The research demonstrated significant differences based on several factors:

• Examiners trained in numerical scoring techniques showed significantly higher accuracy (97.1% versus 86.9%) compared to those without such training, with the advantage strongest among examiners who both learned and applied numerical methods (98.9% versus 88.5%)
• Education level showed complex relationships with accuracy—detection rates did not differ by education level, but false positive errors occurred predominantly among subjects with college degrees
• Horvath noted that blind reviewers lacked access to "special charts" administered in 32 percent of cases (primarily to deceptive subjects), and without these, accuracy appeared lower than the 100 percent accuracy Barland reported for original examiners

The study's findings on verification methods proved particularly influential. Critics excluded some results because confessions were not independent of polygraph examinations, arguing studies were "merely demonstrations that examiners score charts in a similar way" and thus estimates of reliability rather than validity.

Discussion & Significance

Barland and Raskin's 1976 work fundamentally transformed how the scientific community approached polygraph validation. By systematically examining real criminal cases with known outcomes, the research provided empirical data that moved debates beyond anecdotal evidence and laboratory analogues. The study's sophisticated use of blind evaluation and multiple verification criteria set methodological standards that influenced subsequent field research for decades.

The finding that numerical scoring substantially improved accuracy had profound practical implications. The 97.1% accuracy for numerically-trained examiners suggested that systematic, standardized scoring methods—rather than purely clinical judgment—provided measurable advantages, though questions remained about whether superior training or examiner selection explained the differences. This discovery accelerated the profession's movement toward standardized numerical evaluation systems.

However, the study also exposed critical limitations in field validation methodology. The lower accuracy rates compared to some other studies may have reflected that police records of verification had greater variability or were less reliable than those from polygraph firms. These findings sparked ongoing debates about appropriate verification criteria, base rates, and the challenges of establishing true ground truth in operational settings where multiple factors influence case outcomes.

Limitations & Considerations

A fundamental limitation was the dependence on confessions for verification, which critics argued were not independent of polygraph examinations, making results estimates of reliability rather than true validity. This circularity problem—where the polygraph might have influenced the confession used to verify its accuracy—remained a persistent challenge in field validation research.

Interpersonal expectancy effects posed another significant concern, as an examiner's preexamination decision concerning guilt or innocence could affect question construction or the suspect's psychological state, thereby influencing physiological responses even in blind re-evaluations. Additionally, the absence of "special charts" from blind evaluation—charts administered in 32% of cases primarily to deceptive subjects—created systematic bias in accuracy estimates.

• Verification criteria relied heavily on confessions that may have been influenced by polygraph results
• Interpersonal expectancy effects could not be fully controlled even with blind chart evaluation
• Missing data from special chart administration in substantial proportion of cases
• Selection bias from cases chosen for inclusion based on verification availability

Practical Applications

The most immediate practical impact was the validation of numerical scoring systems. The finding that numerical scoring training correlated with accuracy rates approaching 98.9% when properly applied provided empirical justification for professionalizing polygraph practice through standardized training and evaluation protocols. This evidence supported the development of formal scoring algorithms and quality control procedures that became standard in government polygraph programs.

For consumers and policymakers, the research underscored that polygraph accuracy is not an inherent property of the instrument but depends critically on examiner training, technique standardization, and methodological rigor. The study revealed that accuracy claims must be evaluated in context—considering verification methods, decision criteria, base rates, and whether inconclusive results are included in calculations. These insights remain essential for interpreting polygraph evidence in legal, employment, and security screening contexts today.