A non-standard method for estimating the accuracy of lie detection techniques demonstrated on a self-validating set of field polygraph examinations

Ginton (2013) presented a unique ground-truth-free method for estimating CQT accuracy using paired polygraph examinations from criminal cases where two suspects provided contradictory accounts. Using a combined probabilistic and algebraic model on cases from Israeli police investigations (1981-1982), the study circumvented traditional criterion problems by exploiting the logical constraint that one member of each pair must be deceptive.

The study analyzed paired polygraph examinations from criminal cases where two opposing versions were tested. Blind scoring by examiners, combined with algebraic modeling of concordant (both deceptive/both truthful) and discordant (opposite outcomes) pairs, allowed estimation of accuracy rates without external ground truth criteria.

This field validity study addresses the 'criterion problem' in polygraph research by developing a ground-truth-free mathematical method to estimate CQT accuracy. Using paired criminal examinations where two opposing versions were tested, Ginton's probabilistic-algebraic model extracts accuracy estimates without relying on confessions or external evidence. Results showed 94% sensitivity and 83.5% specificity; however, when inconclusive outcomes were included, accuracy dropped to 80% for both groups with 20.3% undecided cases. A subsequent replication in China produced comparable findings (83.6% and 82.2%), suggesting the methodology's robustness. The study is considered a breakthrough for addressing inherent weaknesses in traditional field validity approaches.

This methodology provides a rigorous field-based approach to validating CQT accuracy that circumvents reliance on potentially contaminated external criteria, offering practical value for evaluating polygraph performance in operational settings. The replicability of results across different countries and larger samples strengthens confidence in the method's applicability for future polygraph validity research and policy decisions.

Background & Context

The Comparison Question Test (CQT) has been controversial within the scientific community for many years, representing the most widespread polygraph technique. A persistent challenge in polygraph research has been the "criterion problem"—the difficulty of establishing ground truth in field studies without relying on potentially contaminated evidence like confessions or judicial outcomes.

Ginton's innovative method has been considered a breakthrough by some researchers (Raskin & Kircher, 2014), as it addresses fundamental weaknesses in traditional field validity studies. Due to the unique developed methodology, this research was not subjected to the weaknesses usually found in other polygraph field validity studies.

Ginton developed this approach based on data collected during his tenure heading all polygraph activities in the Israel National Police during 1981-1982, including field labs, a polygraph school, and a research unit. This study represents a novel mathematical solution to estimate CQT accuracy without requiring external verification of truth or deception.

Research Design & Methodology

The study is built on paired examinations in criminal cases in which two opposing versions per case have been subjected to polygraph tests. The developed model is ground-truth free, thus there was no need to rely on external criteria of deception (e.g., confessions or physical evidence) in estimating the accuracy of the CQT.

The methodology employed a combined probabilistic and algebraic model. Key design elements included:

• The examiners were completely blind as to the cases involved, nor did they identify the pairs
• The background of each case was carefully examined by two experienced examiners to eliminate cases in which it was possible that the two parties might have told their subjective truth or that both were lying
• Assuming that in each pair there must be one deceptive and one truthful subject, each pair scored as either two deceptive or two truthful subjects must include an error, while a pair scored as one deceptive and one truthful means either two hits or two errors
• Given the percentage of pairs of each of the three possible combinations of outcomes found in the sample, one can use algebraic equations to extract the values indicating the accuracy and the error rates of CQT

The kinds of crimes or issues involved in the paired examinations cases covered a vast range, ensuring representativeness of typical criminal investigations.

Results & Key Findings

Results indicate an accuracy rate of 0.94 in detecting guilty examinees (Sensitivity) with a 0.06 False Negative rate and an accuracy rate of 0.835 (Specificity) with False Positive of 0.165 for the innocents.

When inconclusive outcomes were factored in, the performance changed substantially:

• These figures excluded 20% of the cases that were ruled inconclusive
• When no inconclusive calls were allowed, the accuracy rate dropped down to 0.8 with 0.2 error rates for both the guilty and the innocent examinees
• The computed accuracy rates of decisions in field examinations were 94% for deceptive and 83.5% for truthful examinees with 20.3% undecided tests

A subsequent replication study in China by Mao, Liang, and Hu (2015) used Ginton's methodology on a larger sample and found similar results, with an accuracy rate of detecting guilty examinees at 0.836 with a 0.164 False Negative rate, and detecting innocents at 0.822 with a 0.178 False Positive rate.

Discussion & Significance

The importance of this research stems from its being a field study that due to the unique methodology was not subjected to weaknesses usually found in polygraph field validity studies. The paired-examination approach circumvents the criterion problem by exploiting the logical constraint that in cases where two suspects provide directly contradictory accounts, one must be deceptive and one truthful.

However, the study has generated substantial scientific controversy. While critics applauded Ginton for his innovative method, some noted how its empirical application was flawed and resulted in inflated estimates of CQT validity. In a recent review by Iacono and Ben-Shakhar (2019), the authors dedicated a significant portion to scrutinizing this novel approach, criticizing its outcomes to the point that nullified any contributions it might have had in dealing with the long-lasting controversy regarding CQT validity.

The method's applicability extends beyond polygraph testing. This method is applicable to other techniques of deception detection and with some necessary adaptations may also be applied to eyewitness situations.

Limitations & Considerations

Several limitations warrant careful consideration:

• The data are based on criminal investigations conducted by the Israeli police over 3 decades ago, and CQTs conducted in this setting may be contaminated
• By the very fact that in paired-tests cases, each examinee faced a person who provided an uncompromised opposite version, this might affect the type of people who take the test by screening out deceptive subjects who cannot confront such a direct accusation, and in case of truthful examinees to increase their anxiety
• The assumption that exactly one member of each pair must be deceptive may not hold universally—some cases might involve mutual misunderstanding or both parties believing their subjective truth
• The method assumes this is the only problem facing CQT field studies, when other validity issues may also be present

The high inconclusive rate (20%) raises questions about practical utility, and the substantial drop in accuracy when inconclusives are forced to decisions suggests that examiner uncertainty plays an important protective role against errors.

Practical Applications

For polygraph practitioners, this research offers both encouragement and caution. The methodology demonstrates that under specific conditions—paired suspects with directly contradictory accounts—CQT field accuracy can be estimated without waiting for external verification. This could provide faster feedback for quality assurance and training purposes.

However, the practical accuracy figures—80% when inconclusives are eliminated—underscore the importance of allowing examiners to render inconclusive decisions rather than forcing binary outcomes. For consumers and policymakers, these results suggest that while the CQT shows better-than-chance performance in field settings, the error rates remain substantial enough to warrant caution in high-stakes decisions, particularly the 16.5% false positive rate for innocent examinees identified in this study.