Polygraph lie detectors are accurate
Polygraphs measure physiological stress responses that have many causes beyond deception, making them unreliable. The definitive 2003 National Academies report found polygraph screening produces unacceptably high rates of false positives and false negatives.
What we know
The polygraph measures changes in breathing, heart rate, blood pressure, and skin conductance on the assumption that lying provokes a distinctive physiological response. This assumption is scientifically weak: the same responses are produced by anxiety, embarrassment, surprise, and many other emotions unrelated to deception. A person who is nervous about being tested may appear deceptive, while a practiced deceiver may remain calm.
In 2003, the National Academies of Sciences published a landmark review, 'The Polygraph and Lie Detection', examining more than a century of research. The review found that while polygraph tests can distinguish lying from truthfulness at rates better than chance in specific-incident investigations, the accuracy drops substantially in screening contexts where subjects face vague, generic questions. The report concluded that polygraph testing yields an 'unacceptable choice' for security screening between too many innocent employees falsely judged deceptive and too many genuine security threats going undetected.
The problem of countermeasures further undermines polygraph reliability. Both cognitive strategies (such as mentally performing arithmetic during control questions) and physical techniques (such as biting the tongue) can, under laboratory conditions, help a deceptive person appear truthful. The NAS noted that individuals with strong motivation and resources, precisely those who pose the greatest security risk, are most likely to learn and use such countermeasures.
Despite this evidence, polygraph use persists in some law enforcement and national security contexts, partly because the belief in its accuracy creates a deterrence effect and elicits confessions from suspects who fear the test. Courts in most US federal jurisdictions and many countries do not accept polygraph results as evidence precisely because of their scientific unreliability.
Common claims
- Polygraphs reliably detect lying with 90%+ accuracy.False. The NAS found accuracy well below perfection, especially for screening.
- Countermeasures cannot fool a polygraph.False. Physical and cognitive countermeasures can defeat the test under lab conditions.
- Passing a polygraph proves innocence.False. Innocent people regularly fail, and guilty people regularly pass.