The Role of Comparison Questions in Physiological Detection of Deception

Experimental study with 120 participants examining the role of comparison questions in producing differential physiological patterns for innocent versus guilty subjects.

Mock crime experiment with 120 participants randomly assigned to guilty and innocent conditions. Physiological responses to relevant and comparison questions analyzed for differential patterns.

Horowitz, Kircher, Honts, and Raskin designed a carefully controlled mock crime study to test the fundamental mechanism underlying CQT effectiveness. With 120 participants, the study examined whether comparison questions produce genuinely different physiological patterns for innocent versus guilty subjects. The finding that they do provided crucial empirical support for the CQT rationale — validating the core assumption that truthful individuals will react more strongly to comparison questions while deceptive individuals will react more strongly to relevant questions.

This study validates the core theoretical mechanism of the CQT. The demonstration of differential response patterns in a controlled experiment with adequate sample size strengthens the scientific foundation for comparison question testing.

Background & Context

The Comparison Question Test (CQT) has been the cornerstone of forensic polygraph practice for decades, yet its underlying theoretical mechanism had remained inadequately tested. The study examined 60 guilty and 60 innocent participants in a mock crime experiment, using different types of comparison questions in four conditions. The fundamental question was whether comparison questions actually produce the differential physiological patterns that polygraph theory predicts—a critical assumption that had been accepted largely on faith rather than empirical validation.

Prior to this research, the field lacked rigorous experimental evidence demonstrating that comparison questions serve their intended purpose. The CQT rationale assumes that innocent examinees will react more strongly to comparison questions (addressing past misdeeds) while guilty examinees focus their concern on the crime-relevant questions. Testing this mechanism directly required a carefully controlled experimental design that could isolate the effects of different comparison question types.

Research Design & Methodology

The study employed 120 participants (60 guilty and 60 innocent) in a mock crime experiment with four distinct conditions testing different types of comparison questions. This design allowed researchers to systematically examine how manipulating comparison question format affects physiological response patterns.

The four experimental conditions were:

• Relevant-Irrelevant (R-I): Participants answered only relevant and neutral questions without comparison questions
• Trivial Directed Lie: Participants instructed to lie to three of six neutral questions
• Personal Directed Lie: Participants instructed to lie to personally relevant questions
• Probable Lie: Traditional probable lie comparison questions were used

Respiration, cardiovascular, vasomotor, and electrodermal activity were recorded, providing comprehensive physiological data across multiple channels. This multi-measure approach strengthened the study's ability to detect differential response patterns and avoided reliance on any single physiological indicator.

Results & Key Findings

Manipulation of the comparison questions produced different patterns of physiological responses for innocent but not for guilty participants. This finding represents the critical empirical validation of CQT theory—innocent examinees showed differential reactivity based on question type, exactly as the theory predicts.

Key findings included:

• Innocent participants showed significantly different physiological patterns across the four comparison question conditions
• Guilty participants maintained consistent response patterns regardless of comparison question type
• The R-I test produced an unacceptable rate of false positive decisions, confirming that comparison questions are essential for accurate discrimination
• Both directed lie variants and probable lie comparison questions produced the theoretically predicted differential patterns

The study's 120-participant sample provided adequate statistical power to detect meaningful differences across conditions. Based on a doctoral dissertation submitted to the University of Utah by S. W. Horowitz, a partial report of these results was presented at the Society for Psychophysiological Research in San Francisco in 1988, indicating the work underwent peer scrutiny before publication.

Discussion & Significance

This study provided the first rigorous experimental demonstration that comparison questions function as theorized in CQT examinations. The finding that innocent examinees show variable physiological responses depending on comparison question type—while guilty examinees do not—validates the core psychological mechanism underlying comparison question testing. This addresses a fundamental criticism that the CQT lacked empirical support for its theoretical foundation.

The practical implications are substantial. The study demonstrated that the relevant-irrelevant format without comparison questions produces unacceptably high false positive rates, confirming that comparison questions serve an essential function rather than being merely traditional practice. Furthermore, the finding that multiple comparison question formats (directed lie and probable lie) produce appropriate differential patterns suggests flexibility in test construction while maintaining scientific validity.

The research has been influential in subsequent polygraph literature, with reviews consistently citing it as key evidence supporting CQT validity. The study strengthened the scientific foundation for comparison question testing at a time when polygraph techniques faced increasing scrutiny regarding their theoretical and empirical basis.

Limitations & Considerations

As with all laboratory mock crime studies, the ecological validity requires careful consideration. Participants faced substantially lower stakes than real-world examinees, potentially affecting both motivation and physiological arousal patterns. The "guilty" participants committed simulated rather than actual crimes, and consequences for test failure were minimal compared to criminal investigations.

The study examined physiological response patterns but did not report final accuracy rates for classifications, focusing instead on the underlying mechanism. While this addresses the theoretical question, practitioners and policymakers often seek concrete accuracy estimates. Additionally, the four-condition design, while methodologically sophisticated, meant smaller sample sizes within each condition compared to the overall 120-participant pool.

Practical Applications

For polygraph examiners, this research validates the critical importance of properly formulated comparison questions. The unacceptable false positive rate with the R-I format demonstrates that comparison questions are not optional refinements but essential components. Examiners can draw confidence from evidence that both probable lie and directed lie comparison questions produce the theoretically expected differential patterns, allowing appropriate selection based on case requirements and examinee characteristics.

For consumers and policymakers evaluating polygraph evidence, this study demonstrates that CQT has moved beyond assertion to empirical validation of its core mechanism. The research shows that when properly administered with appropriate comparison questions, the test operates according to established psychophysiological principles rather than examiner intuition. However, the study also underscores that test format matters significantly—not all questioning approaches provide equal validity.