Polygraph Field Test Accuracy of CQT in Criminal Cases

Field validation using confirmed criminal case outcomes achieving accuracy rates above 90% for both truthful and deceptive classifications.

Field study using criminal cases with confirmed outcomes through confessions and independent evidence.

Published in 1987 in Law and Human Behavior, 11, 307–319, this work by Driscoll and colleagues examined polygraph field test accuracy of cqt in criminal cases. Key findings include: Field validation using confirmed criminal case outcomes. Achieved accuracy rates above 90% for both truthful and deceptive classifications, consistent with laboratory findings. As a field study conducted with real cases, this research provides ecological validity that laboratory studies cannot match.

Field-laboratory consistency in accuracy rates supports the generalizability of controlled research to operational settings.

Background & Context

During the mid-to-late 1980s, polygraph research faced a critical challenge: demonstrating that laboratory findings could translate to real-world criminal investigations. The 1983 U.S. Office of Technology Assessment had raised serious questions about polygraph validity, creating an urgent need for rigorous field studies using actual criminal cases with confirmed outcomes.

This research emerged from a larger National Institute of Justice-funded project conducted by David Raskin, John Kircher, Charles Honts, and colleagues at the University of Utah. Their work represented one of the most comprehensive attempts to validate the Control Question Technique (CQT) using real criminal cases from federal law enforcement. The study aimed to bridge the gap between controlled laboratory experiments and the messy reality of criminal investigations.

The timing was particularly significant. Federal agencies were expanding polygraph use, and courts were grappling with admissibility standards. The research sought to demonstrate that "the accuracy of the CQT can exceed 90% when properly employed by skilled examiners"—a claim that would either validate decades of practice or expose fundamental flaws in forensic polygraph application.

Research Design & Methodology

The study utilized 241 confirmed and unconfirmed polygraph charts from examinations conducted by the U.S. Secret Service during FY 1983 and FY 1984, drawn from a larger database of over 1,700 criminal investigations. This represented genuine field testing under operational conditions, not laboratory simulations. The cases involved specific-incident criminal investigations where suspects were examined using standard CQT procedures.

Cases were obtained from federal law enforcement agency files, with confirmations based on confessions that were supported by evidence. This ground truth criterion was essential—without independent confirmation of guilt or innocence, accuracy calculations would be meaningless. The research team employed blind evaluation procedures where multiple experienced polygraph examiners independently scored the charts without knowledge of case outcomes.

The methodology included several innovative elements:

• Assessment of polygraph examiners with different educational backgrounds and experience levels
• Evaluation of computer-based interpretation methods developed at the University of Utah
• Analysis of the extent to which laboratory mock-crime experiments provide results applicable to field settings
• Use of standardized numerical scoring systems to quantify physiological responses

Results & Key Findings

The field validation study produced remarkably strong accuracy outcomes that supported the operational use of polygraph testing in criminal investigations. Original examiners' decisions on individual relevant questions ranged from 91-96% correct on confirmed truthful and 85-95% correct on confirmed deceptive answers, demonstrating that polygraph testing in real-world settings could achieve the high accuracy rates previously observed primarily in laboratory research.

When charts were blindly re-scored by independent examiners unfamiliar with the cases, accuracy remained substantial though somewhat lower. Blind interpretation produced accuracies from 63-85% on truthful and 84-94% on deceptive answers. This difference between original and blind scoring revealed the value of contextual information and examiner-subject interaction during testing, while also demonstrating that physiological data alone retained significant diagnostic value.

Computer-based interpretation methods showed exceptional promise:

• Computer interpretation accuracy ranged from 95-96% on confirmed truthful responses
• Computerized methods outperformed blind human interpretation in most comparisons
• Algorithms developed at the University of Utah demonstrated the feasibility of objective, automated decision-making
• The accuracy of human and computer interpretations was very high overall

Importantly, the results showed substantial similarity between data sets obtained from laboratory subjects and criminal suspects when comparable procedures were used, providing critical evidence for the generalizability of experimental polygraph research to operational settings.

Discussion & Significance

This research provided some of the strongest empirical support for CQT validity in forensic applications up to that time. The above-90% accuracy rates for both truthful and deceptive determinations suggested that when properly administered by trained examiners, polygraph testing could serve as a valuable investigative tool. The convergence between laboratory and field data was particularly important, validating decades of experimental research.

The superior performance of computer algorithms pointed toward a future of more standardized, objective polygraph interpretation. The findings suggested that available computer methods could be used effectively to increase the objectivity and accuracy of credibility assessment of criminal suspects. This represented a potential solution to long-standing criticisms about subjective examiner judgment and lack of standardization across the profession.

However, the research also revealed important nuances. The findings indicated that adjustments in decision criteria should be made to reduce the false positive errors caused by overprediction of deceptive field outcomes using criteria derived from laboratory experiments. This suggested that direct application of lab-derived decision rules to field settings could inappropriately classify truthful individuals as deceptive—a finding with serious implications for innocent suspects.

Limitations & Considerations

Several methodological constraints tempered the study's conclusions. The research relied on confessions supported by evidence as ground truth, but confessions themselves can be unreliable, particularly if obtained through coercive interrogation following failed polygraph examinations. This circular reasoning problem—where polygraph results might influence confession likelihood—could inflate apparent accuracy rates.

The study examined only federal law enforcement cases, primarily from the Secret Service, which may not represent the broader landscape of polygraph practice. Federal examiners typically receive extensive training and work within standardized protocols. Many challenges facing polygraph arise from poor training and lack of standardization in the polygraph profession, suggesting these high accuracy rates might not generalize to less rigorous operational settings.

Additionally, the research could not control for pre-selection effects. Cases selected for polygraph examination may differ systematically from the general suspect population, potentially in ways that make deception easier to detect. The study also could not address potentially effective countermeasures that sophisticated examinees might employ to defeat the test.

Practical Applications

For practitioners, this research validated the CQT's use in specific-incident criminal investigations when conducted according to established protocols. The findings supported continued use of polygraph examinations as investigative tools, though not necessarily as definitive proof of guilt or innocence. The 85-96% accuracy range across different scoring methods suggested polygraph results could appropriately guide investigations while requiring corroboration through traditional evidence.

For consumers and defendants, the research provides both reassurance and caution. While properly conducted polygraph examinations by qualified examiners can achieve high accuracy, the variability between original and blind scoring underscores the importance of examiner quality and standardized procedures. Anyone facing a polygraph examination should ensure the examiner is properly credentialed and that testing follows established professional standards rather than idiosyncratic methods.