A Validation and Reliability Study of Counterintelligence Screening Tests

US military study of polygraph screening test validity for counterintelligence applications.

Military field study validating counterintelligence screening polygraph test format.

Barland's military study validated polygraph screening for counterintelligence, finding detection rates significantly better than chance though with higher inconclusive rates than specific-issue tests. The screening context — where base rates of deception are typically low — presents the most challenging testing scenario.

Better-than-chance detection in low-base-rate screening confirms some utility for security applications, though the higher inconclusive rate suggests screening tests require different interpretation frameworks than specific-issue tests.

Background & Context

The Counterintelligence Screening Test (CIST) emerged in 1971 as a significant innovation in military polygraph screening, developed by U.S. Army Military Intelligence examiners at Fort Meade. The CIST introduced a controversial change by using the directed lie control question (DLCQ) as a criterion for evaluating a subject's level of reactivity, departing from the traditional probable lie control questions used in most polygraph formats at the time.

Neither the validity of the DLCQ nor the validity of the CIST format had been established under controlled conditions when this study was undertaken to determine whether the CIST could accurately differentiate between truthful and deceptive subjects in a mock screening situation. This represented a critical research gap during the height of Cold War counterintelligence operations, when thousands of military and intelligence personnel required security screening.

The study addressed a fundamental challenge in polygraph science: validating screening tests in environments where the base rate of deception is typically extremely low. Unlike specific-issue criminal investigations where there's often strong evidence suggesting deception, counterintelligence screening examines individuals with no particular suspicion against them—presenting the most demanding test scenario for polygraph accuracy.

Research Design & Methodology

The study involved 56 subjects who were given CIST polygraph examinations concerning five relevant questions, with ground truth previously established by background investigation. This design allowed for objective verification of accuracy, as examiners' determinations could be compared against known facts about each participant.

The 56 subjects were randomly assigned to two groups: truthful subjects (n=26) who were instructed to answer all five relevant questions truthfully, and deceptive subjects (n=30) who were instructed to falsify their answer to one randomly selected relevant question while answering the other four truthfully, with deceptive subjects offered $20 if they could beat the polygraph. This monetary incentive created motivation to defeat the test, simulating real-world counterintelligence threats.

The study compared three different methods of evaluating polygraph charts: zone comparison, greatest control, and relevant-irrelevant approaches. This methodological comparison allowed researchers to determine which scoring system produced the most accurate results with the CIST format. The polygraph examiner's task was to determine whether each subject was truthful or deceptive, and if deceptive, to which question(s).

Results & Key Findings

The study demonstrated that the CIST technique could differentiate between truthful and deceptive subjects at statistically significant levels. Using the zone comparison scoring system, the examiners correctly categorized 37 (66 percent) of the 56 subjects, made no decision in 9 (16 percent) of the cases, and erroneously categorized 10 (18 percent) subjects—five false positive and five false negative errors.

When inconclusive cases were excluded from the analysis, accuracy improved substantially. Excluding the nine inconclusive cases, 79 percent of the examiners' decisions were correct (p less than .001), demonstrating accuracy significantly better than chance. This statistical significance indicated the CIST technique had genuine discriminatory power beyond random guessing.

The comparative analysis of scoring methods revealed important differences:

• Zone comparison method: Successfully identified both truthful and deceptive subjects at better than chance levels, producing the most balanced accuracy
• Relevant-irrelevant method: Also identified both truthful and deceptive subjects effectively, performing comparably to zone comparison
• The greatest control method was able to identify truthful subjects but could only identify deceptive subjects at chance levels

The results indicated that the CIST technique, incorporating directed lie control questions, was able to differentiate between truthful and deceptive subjects, though the greatest control method was inferior to the zone comparison and relevant-irrelevant methods in its inability to identify deceptive subjects. The error distribution showed equal rates of false positives and false negatives, suggesting no systematic bias toward either type of error.

Discussion & Significance

This study provided the first controlled validation of directed lie control questions in a screening context, establishing a scientific foundation for a technique that would become standard in U.S. government counterintelligence operations. The finding that the CIST performed significantly better than chance validated its continued use, though the 16% inconclusive rate highlighted inherent challenges in screening scenarios where base rates of deception are low.

The study's confirmation that zone comparison and relevant-irrelevant methods outperformed the greatest control approach had immediate practical implications for field examiners. It demonstrated that not all scoring systems work equally well with directed lie questions, and that the choice of chart interpretation method significantly impacts accuracy. This finding influenced subsequent development of polygraph screening protocols across military and intelligence agencies.

The balanced error rate—equal numbers of false positives and false negatives—represented an important characteristic for security screening applications. Unlike specific-issue testing where false positives may be more tolerable, counterintelligence screening requires balanced accuracy to avoid both security risks (false negatives) and unfair career consequences for innocent personnel (false positives). The study demonstrated the CIST achieved this balance while maintaining better-than-chance overall accuracy.

Limitations & Considerations

The mock crime scenario, while methodologically necessary, created potential limitations in generalizability to actual counterintelligence threats. Additional research was needed to assess the accuracy of the CIST technique relative to other polygraph screening formats. The psychological stakes of losing $20 in a laboratory setting likely differ substantially from the consequences facing actual spies or individuals with genuine security violations.

The study's design—where deceptive subjects lied to only one of five relevant questions while answering four truthfully—may not fully represent real-world counterintelligence scenarios where subjects might have multiple areas of deception. Additionally, the relatively small sample size (56 subjects) limited the precision of accuracy estimates, particularly for analyzing error rates in specific subgroups. The study also could not address the deterrent effect that periodic screening might have on potential security violations.

Practical Applications

This research validated the operational use of the CIST format with directed lie control questions for military counterintelligence screening, establishing accuracy benchmarks that informed expectations for field applications. The finding that zone comparison and relevant-irrelevant scoring methods produced superior results provided concrete guidance to examiners selecting chart interpretation approaches. Polygraph programs could use the documented 79% accuracy rate (excluding inconclusives) and 16% inconclusive rate to set realistic performance standards.

For security managers and policymakers, the study confirmed that polygraph screening provides information gain beyond chance while acknowledging the inevitable presence of inconclusive results and occasional errors. The balanced false positive and false negative rates suggested that screening programs should incorporate procedures for investigating inconclusive and deceptive outcomes rather than relying solely on polygraph results for security decisions. This research established that counterintelligence screening, while imperfect, offers better-than-chance detection when properly administered using validated techniques and scoring methods.