Do Lie Detectors Really Catch Lies?

Syed Aftab Shah explains how lie detector tests work and how accurate their findings are. According to the author, the polygraph tests are misunderstood, as they don't tell if someone's lying or not, but only record the physiological responses to the questions.

Aldrich Aimes, a former Central Intelligence Agency (CIA) officer was allegedly paid $1.5 million by the Russians to pass on sensitive information about US intelligence operations until his arrest in 1994. As a double agent, he played a cat and mouse espionage game for nine years and successfully managed to cheat two lie detector tests administered by the CIA.

In 2000, he penned down a letter to Dr. Steven Aftergood of the Federation of American Scientists about the reliability of lie detector tests. The letter starts with the following excerpt:

“Dear Mr. Aftergood,

Having had considerable experience with the polygraph (well beyond that which you referred to), I read your very sensible essay in Science with great interest. I offer you a few comments on the topic for whatever interest or use they may have.

Like most junk science that just won't die (graphology, astrology and homeopathy come to mind), because of the usefulness or profit their practitioners enjoy, the polygraph stays with us.”

In the investigative process, when a suspect is politely called to the police station to give their statement, most people will be willing to cooperate. Maybe because there is nothing to hide if you are innocent, but mainly it’s a matter of impression as non-cooperation leads to more suspicion. The person in charge of questioning the suspect is tasked with an important challenge of detecting deception through verbal, and non-verbal cues. Detecting verbal deception is given precedent over non-verbal deception, as the former can easily translate into a confession. For instance, a suspect who is lying may not be able to keep a track of their own story due to memory load (Vrij, Leal, Mann & Fisher, 2012). In contrast, identifying deceit through non-verbal cues is very tricky and only a handful of cues including, constant fidgeting, or avoiding eye contact, have been consistently associated with lying (Kraut, 1980).

The polygraph test, popularly known as a lie detector test, is often misunderstood by the general public due to its portrayal in the mainstream media. The test does not indicate if someone is lying or telling the truth. In reality, it measures only physiological responses to arousals in the form of elevated heart rate, breathing and sweating of fingertips (The Polygraph and Lie Detection, NRC, 2003).

There are two main techniques of polygraph tests, the Control Question Test (CQT) and the Concealed Information Test (CIT) also known as the Guilty Knowledge Test (GKT).

CQT is a widely used approach, particularly in the US, Canada, and Israel, with three categories of questions: 1) neutral questions, 2) control questions, 3) relevant questions. Neutral questions are common knowledge. The second category includes control questions, which are probable lies. The answers to these questions are already known by the examiner, and they are intended to set a baseline level for the polygraph machine. The third category is the most crucial part of the interview, as questions related to the investigation.

The person who is being deceptive will show higher levels of physiological arousal in relevant questions as compared to control questions, and the arousal response will be vice versa in the case of a person being honest. To elaborate further, a control question of “have you ever stolen something in your life?”, is implanted to create a dilemma, as the answer will almost always be a probable lie. A person who is innocent is assumed to be lying in this question with a higher level of arousal, as compared to telling the truth in the relevant question “did you murder that man?”. In contrast, a person who is being deceptive is assumed to be lying in both control and relevant questions, while having somewhat of a similar arousal level.

These assumptions about arousal levels of the examinee, be it innocent or guilty, are based on these theories: Threat of Punishment Theory (Davis, 1961), Related Arousal Theory (Prokasy & Raskin, 1973), Dichotomization Theory (Ben–Shakhar, 1977) and Psychological Set Theory (Barland, 1981).

A study conducted by Professor Don Grubin in 2010, found the accuracy of CQT between 74% to 89% for guilty examinees with a false negative error range of 1% to 13%, whereas the accuracy is at a range of 59% to 83% for innocent examinees with a false positive error range of 10% to 23%.

The National Research Council’s review called ‘The Polygraph and Lie Detection’ (2003) took up 37 laboratory studies and 7 field studies for the purpose of evaluating the accuracy of the CQT technique, and estimated an average accuracy of 85%. The review went on to criticize the failure of polygraph test studies to make use of conceptual, theoretical, and technological advancement in science, that are specifically relevant to the psychophysiological detection of deception:

“Notwithstanding the limitations of the quality of the empirical research and the limited ability to generalize to real-world settings, we conclude that in populations of examinees such as those represented in the polygraph research literature, untrained in countermeasures, specific-incident polygraph tests can discriminate lying from truth-telling at rates well above chance, though well below perfection. Because the studies of acceptable quality all focus on specific incidents, generalization from them to uses for screening is not justified.” (National Research Council, 2003, p. 4)

After his incarceration, Aldrich Ames was interviewed by the Daily Press and when he was asked how he managed to cheat the machine twice, he said: ''Confidence is what does it. Confidence and a friendly relationship with the examiner… rapport, where you smile and you make him think that you like him.'' If the purpose of the polygraph test is to gain a confession, then inducing fear and anxiety of being caught lying makes all the difference. The portrayal of polygraph tests as lie detectors in the mainstream media has instilled a perception of just that – a machine that catches all your lies.

Case in point, an experiment demonstrated that those participants who were ‘wired up’ to a fake ‘lie detector’ machine were more likely to admit to embarrassing beliefs and facts, as compared to participants who weren’t ‘wired up’ at all. This is formally known as the ‘bogus pipeline’ effect (Jones & Sigall, 1971; Tourangeau, Smith, & Rasinski, 1997).

A lesser-used technique of polygraph test is the Concealed Information Test (CIT), in which specific information guised in a ‘critical’ question is planted within the polygraph test. This information can only be known by the perpetrator of the crime, for instance the type of weapon used to murder someone. A person being deceitful will show higher arousal level when the concealed information is revealed. The GKT takes approximately 90 to 120 minutes to administer and usually contains about 5 questions. Each of these questions are repeated three times with different sequences of alternatives. (Hira & Furumitsu, 2002).

CIT technique is part of the investigative procedure of law enforcement in Japan since 1950s. With more than 5,000 polygraph examination conducted each year since 1970s, the result revealed an accuracy rate of spotting deceit at 90 percent (Yamamura & Miyata, 1990). This technique’s theoretical framework is based on Orienting Response Theory, a ‘what is it?’ natural response of the body to novel change in the environment.

In meta-analysis research conducted by Ben-Shakhar and Elaad (2003) with a sample of 80 studies that used CIT/GKT technique, found an average correlation coefficient between guilt and innocence at 0.79. The researchers concluded with a recommendation to introduce GKT technique for criminal investigative purposes:

“Our meta-analysis demonstrated that the GKT has excellent potential as an applied method for detecting information and for differentiating between individuals possessing guilty knowledge and innocents. Furthermore, the GKT has various advantages over alternative psychophysiological detection methods because it is based on sound theoretical foundations and a standardized procedure. This raises a question regarding the limited usage of the GKT in criminal investigations in North America.” page 38.

To leave you with one last statistic. A survey was conducted by Iacono and Lykken (1997) to gauge the academic opinion of the Society for Psychological Research (SPR), and the American Psychological Association’s (APA) Division 1, regarding their confidence in polygraph test techniques i.e. CIT/GKT versus CQT. The result showed that 77% of SPR and 72% of APA Division 1 believed CIT to be based on scientific principles. In comparison to CQT, the result was 36% and 30% respectively.

Nevertheless, the researchers of the above survey made it clear that a vote cannot decide which type of technique is better. The CQT is used as a screening tool for recruitment and interrogative purposes by the government, as well as private organizations. In contrast, the CIT is used for investigative purposes with a prerequisite of having concealed information related to the crime. Finally, a machine that catches a human lie with full accuracy has not been invented yet, at best polygraph tests have the potential to be an effective method in aiding investigators in detecting deception and narrowing the suspect list.