Lie Detection

Relevant-Irrelevant Technique

Modern polygraph testing is computerized and involves obtaining digital representations of autonomic nervous system activity associated with palmar sweating, respiration, and blood pressure while individuals are asked different types of questions. The first broadly used polygraph techniques became known as relevant-irrelevant tests (RIT). The RIT includes a relevant or “did you do it” question (e.g., in the case of a woman's rape, “Did you place your finger in Glenda's vagina?”). The physiological reaction to relevant questions is compared to the response to truthfully answered irrelevant questions dealing with simple facts (e.g., “Is your name John?”). Guilt is inferred if the relevant question elicits the stronger response. The problem with the RIT is that the accusatory relevant question can be emotionally arousing even when answered truthfully, thus ensuring that it will elicit a strong physiological reaction, often much stronger than the response to the innocuous irrelevant question. Recognized for this bias against the innocent, the RIT has been largely replaced by the control or comparison question test (CQT).

Control (Comparison) Question Technique

The CQT, first introduced in the 1940s, is typically used in forensic applications. Like the RIT, the CQT includes relevant questions. However, the reactions to the relevant questions are compared to those from “control” questions that are thematically related to the content to the relevant question. Control questions, which are answered “no,” are believed to tap a “probable lie” based on the expectation that people typically engage in misbehaviors that are covered by the question. An example of a control question appropriate for an interrogation concerning a sex crime would be “Have you ever committed a sex act you were ashamed of?” CQT theory assumes that innocent individuals will be disturbed more by their denial to the control than to the relevant question because only the control question is likely to elicit a lie. By contrast, guilty individuals are expected to respond more to the relevant question because, for them, this question elicits a more important lie.

Regrettably, for the innocent, the significance of the accusation contained in the relevant and control questions is not equivalent. Only the relevant question deals with the criminal allegation, and only the response to this question can lead to consequences such as prosecution or public embarrassment. Just as is the case for the RIT, the CQT is biased against the innocent because only the relevant question deals with a consequential allegation. Hence, the accuracy [Page 471]of failed CQTs is suspect. But can passed tests be trusted? Unfortunately, research has shown that liars can augment their physiological reactions to the control questions by lightly biting their tongue or doing stressful mental exercises, thereby beating a CQT. Neither failed nor passed CQTs can be trusted.

...