Positions of truthfully answered controls on control question tests with the polygraph

Dans le cadre d'une simulation de crime, des etudiants (N = 120) classes soit comme coupables, comme innocents et informes des details du crime ou innocents et non informes des details du crime, ont ete soumis au detecteur de mensonge a l'aide d'une version modifiee du Control Question Test (CQT). Les questions de controle ambigues, susceptibles d'entrainer un mensonge, ont ete transformees en questions claires et directes qui appelaient des reponses de bonne foi. Lorsque ces questions de controle etaient posees avant les questions reliees au crime, la plupart des participants coupables et innocents etaient classes correctement. La majorite des participants ont echoue au test du polygraphe lorsque les questions reliees au crime leur etaient posees avant les questions de controle. Lorsqu'un participant mentait a une question reliee au crime ayant ete posee dans un deuxieme temps, on observait des reactions plus importantes au plan de la resistance de la peau, F(2, 108) = 8.2, et de la volemie, F(2, 108) = 6.1, comparativement a celles observees lors des reponses d'ordre informatives donnees aux questions de controle initiales. L'exactitude de la detection depend donc de la position des questions de controle.

Mock crime study with 120 students divided into guilty, informed innocent, and uninformed innocent groups. Modified CQT used truthfully-answered control questions with systematic manipulation of question order (controls-first vs. relevants-first). Physiological measures included skin resistance and blood volume.

Background & Context

The Comparison Question Test (CQT) has long been the dominant polygraph technique in criminal investigations, but it rests on a controversial foundation: the use of ambiguous "control" questions designed to elicit lies from examinees. Traditional control questions are intentionally ambiguous and likely to provoke deceptive responses, creating ethical concerns and theoretical challenges for the validity of the test.

Cullen and Bradley's 2004 study challenged this fundamental assumption by testing whether control questions needed to be deceptive at all. Their research examined a radical modification: replacing the typical ambiguous, lie-inducing control questions with clear and direct questions similar to those used in the Concealed Information Test. This represented a significant departure from standard CQT practice and addressed long-standing criticisms about the ethical and logical basis of comparison questions.

The study also investigated a critical yet under-researched element of polygraph testing: question order. The position of control questions relative to relevant questions could theoretically influence physiological arousal patterns, yet this had received minimal empirical attention in the polygraph literature.

Research Design & Methodology

The study employed a mock crime paradigm with 120 university students divided into three groups: guilty participants who committed a simulated crime, innocent participants who were informed about crime details, and innocent participants who were uninformed about the crime. This design allowed researchers to test detection accuracy across different levels of knowledge and involvement.

The critical innovation was the modification of control questions. Instead of using traditional ambiguous questions that examinees would likely answer deceptively (e.g., "Have you ever stolen anything?"), researchers used clear, direct questions answered truthfully, similar to CIT format—for example, if the relevant question was "Did you steal $100?" a control might be "Did you steal $150?"

The experimental manipulation focused on question position: some participants received control questions before relevant questions, while others received relevant questions first. Physiological measures included skin resistance and blood volume, which were analyzed using standard polygraph scoring procedures.

Results & Key Findings

When control questions were presented before crime-relevant questions, most guilty and innocent participants were correctly classified. This represented strong evidence that truthfully-answered control questions could function effectively in a modified CQT format.

However, question order proved critical to test validity:

• The majority of participants failed the polygraph test when crime-relevant questions were asked before control questions
• When participants lied to a crime-relevant question asked second, significantly larger reactions were observed for skin resistance, F(2, 108) = 8.2, and blood volume, F(2, 108) = 6.1
• These physiological responses were compared to truthful informative answers given to initial control questions

The central conclusion was unambiguous: detection accuracy depends on the position of control questions. This finding revealed a previously unrecognized vulnerability in polygraph test construction and challenged assumptions about the robustness of the CQT format.

Discussion & Significance

This research fundamentally challenged the necessity of deceptive control questions in the CQT. The successful classification of participants using truthfully-answered controls demonstrated that the psychological mechanism underlying the CQT may not require examinees to lie to comparison questions—contradicting decades of polygraph theory and practice.

The dramatic effect of question order exposed a critical methodological weakness. If simply reversing the sequence of control and relevant questions could cause most participants to fail the test regardless of actual guilt, this suggested that arousal patterns were influenced by primacy effects, habituation, or strategic psychological factors beyond simple differential salience. This finding aligned with broader critiques of CQT validity and raised concerns about standardization in field practice.

The study provided empirical support for an alternative approach to polygraph testing that could address ethical objections while maintaining detection capability. By demonstrating that clear, truthfully-answered questions could serve as effective controls, the research offered a potential path toward more transparent and defensible polygraph procedures—though adoption of such modifications in operational settings remained uncertain.

Limitations & Considerations

As a laboratory study using mock crimes with student participants, the research faced the typical limitations of analog polygraph research. Real criminal investigations involve substantially higher stakes, greater emotional arousal, and different motivational structures than experimental simulations. The generalizability of these findings to field settings where consequences are genuine remained an empirical question.

The study's small effect on detection accuracy when using the modified format—while statistically significant—required replication with larger samples and diverse populations. Additionally, the specific types of truthfully-answered control questions used may not generalize well to all crime scenarios, and the optimal construction of such questions for different investigative contexts required further research.

Practical Applications

For polygraph examiners, this research highlighted the critical importance of standardizing question order. The findings suggested that protocols should mandate presenting control questions before relevant questions to maximize validity. Any deviation from standardized sequences could artificially inflate false positive rates and undermine test credibility.

The demonstration that truthfully-answered controls can be effective opened possibilities for more ethically defensible polygraph procedures. Examiners could potentially adopt transparent control question formats that don't rely on deception or psychological manipulation, potentially improving the acceptability of polygraph testing in legal and employment contexts. However, such modifications would require careful validation and training before operational implementation.