'Lie Detectors' are Junk Science
Governments and businesses should be banned from using them
Check out more on law at the Evidence Files Medium.
Introduction
People lie. Criminals lie. This has been true for probably all of human existence, as illustrated in the oldest literary works. Unsurprisingly, humans have spent just as long trying to uncover lies through various methodologies. Despite an existence-long penchant for playing fast and loose with the truth, humans remain notoriously bad at detecting lies, with a mere 54% average accuracy. Police and others supposedly trained in detecting deception perform no better than untrained, randomly selected individuals.
Discerning liars from truth tellers is stubbornly hard because there is no formulaic way of doing it. Put differently, people lie in all kinds of ways and they feel differently about whether and when they do it. When caught, some react with guilt, some lie some more, and some just don’t give a damn.
The idea that a simple, universal procedure can be used to determine when a person is lying is almost as old as lying itself. In previous eras (and—sadly—still today), many believed torture would do the trick. In the Industrial Age, when people held budding enthusiasm that machines would eventually handle all sorts of human tasks, inventors started churning out the first mechanized lie detectors, aka polygraphs. Torture and polygraphs as “truth extractors,” however, both rely on junk science. Neither of them show any more effectiveness than plain old guessing. Worse, their unreliability tends to impede more useful approaches to discovering what actually happened, among other harms.
Many police agencies still use the polygraph either in criminal investigations or as part of their employment screening process. People who administer the test are called “polygraph examiners,” a title that gives credence to a pointless, harmful job. An entire organization makes lots of money certifying people for this occupation, articulating its role as follows:
Established in 1966, the American Polygraph Association (APA) is the world's leading association dedicated to the use of evidence-based scientific methods for credibility assessment.
Indeed, the American polygraph industry is worth at least $2 billion. The number globally must be substantially larger.
How do Polygraphs work?
Polygraphs do not “detect lies.” Instead, they measure physiological responses based on questions asked by an examiner. Specifically, they quantify changes in conductivity and skin resistance, blood pressure, heartrate, and respiration. From the measurements of these responses, examiners determine the “likelihood” that someone has answered a question with deception or that they know something about the information presented.
For criminal investigations, examiners use primarily two tests: the Comparison Question Test (CQT) and the Concealed Information Test (CIT).
Comparison Question Test (CQT)
In 2020, Charles R. Honts, Steven Thurber, and Mark Handler conducted a meta-analysis of the most commonly used polygraph test, the CQT. The CQT poses questions to examinees whose answers the examiner rates as Truthful, Inconclusive, or Deceptive. The content of questions fall in two categories: those that relate to the criminal case at issue and so-called comparison questions, the latter of which “usually deal with past probable misdeeds of the examinees that they choose to deny.” Comparison questions tend to focus on personal matters an individual would not want to admit, such as having had committed some kind of morally or socially unacceptable act.
To create a list of questions related to the crime, the examiner analyzes the case file. For the personal questions, the examiner conducts a non-interrogative interview to tease out information that can then be leveraged in the polygraph examination itself. One can already see fundamental flaws in this approach.
Examples of CQT questions, marked as Directly Relevant (DR), Not Necessarily Relevant (NR), or Neutral (NL) to the crime might include:
Were you on Johnson Drive on January first? (DR)
Did you show Victim A a gun? (DR)
Have you ever stolen something? (NR)
Have you ever used a weapon against someone? (NR)
Is today Tuesday? (NL)
The idea behind the CQT is that a criminal would be less worried about revealing certain personal habits than his criminal activity. A person innocent of any crime, on the other hand, would react more strongly to discoveries about their personal conduct because they have no other concerning activity to conceal. Both instances rely on the notion that when under the stress of facing undesired revelations, people exhibit physiological reactions.
Honts, Thurber, and Handler reached the conclusion that the polygraph CQT provides a useful tool in criminal investigation, yet they highlighted numerous flaws. For instance:
[E]thical standards are provided by the various professional groups (e. g., APA, 2015). However, those standards are rarely and inconsistently enforced and have no force over non-members.
They also pointed out that studies supporting the validity of the test are usually conducted by researchers
involved professionally in polygraph testing, either in conducting CQT tests, conducting funded research, or appearing as experts in courts of law supporting or opposed to CQT testing.
In other words, those advocating the use of polygraphs make money from using the device. The authors seem to fail to see the irony in concluding that the polygraph remains a legitimate method for lie detecting given that two of the three of them are subject to precisely the same conflicts:
Charles Honts is licensed as a polygraph examiner and conducts forensic polygraph examinations. He also serves as a consulting and testifying expert witness concerning the quality of polygraph examinations and about the use of polygraph examinations as contributors to false confessions.
Mark Handler is licensed as a polygraph examiner and conducts forensic polygraph examinations. He also serves as a consulting and testifying expert witness concerning the quality of polygraph examinations. The third author is the editor of the journal, “Polygraph & Forensic Credibility Assessment: A Journal of Science and Field Practice.”
Concealed Information Test (CIT)
Of the various types of polygraph examinations, CIT is among the newest. First used widely in Japan, the basic principle behind it is to decide “whether a person recognizes crime-relevant information that innocent people would not know.”
Here is how the FBI explains the way the CIT is conducted:
The body of a victim who apparently died of stab wounds is discovered inside a vacant house. Investigators analyze evidence from the crime scene to identify items most likely to be memorable and important to the perpetrator. The following objects are noted: a broken window used to gain entry; a bloody knife, identified as the murder weapon, near the body; a towel used to gag the victim; duct tape used to bind the victim; and a sleeping bag that covered the body. The investigating agency keeps this holdback information confidential.
During a CIT, the examiner instructs the suspect to answer “no” to all questions on the test. The actual perpetrator of the crime usually has physiological reactions to the correct, or key, items that can be discriminated from those pertaining to the decoy or control items. It is unlikely that someone not involved with the crime would consistently have strong reactions to the key items. With the administration of numerous tests, it becomes almost mathematically impossible for an innocent person to react randomly to the majority of key questions.
This test adopts the same physiological analysis that the CQT does, but makes determinations using a different psychological framework. Both rely heavily on psychological theory, but the American Psychological Association takes umbrage with its value in this sort of application, stating:
There is no evidence that any pattern of physiological reactions is unique to deception. An honest person may be nervous when answering truthfully and a dishonest person may be non-anxious. Also, there are few good studies that validate the ability of polygraph procedures to detect deception.
So, does the polygraph work?
Plainly and simply, no. While polygraphs do effectively detect physiological reactions, no matter the paradigm of questioning, the results cannot provide any useful information to reach conclusions about the veracity of the answers.
The primary problem is that the polygraph “measures peripheral arousal, not deception itself.” Surface level physiological responses vary widely among individuals. Certain of them, like cardiovascular and respiratory responses, can be controlled by the individual, allowing certain people to “beat” the test.
Moreover, while these systems may change autonomously when excited, changes do not provide any evidence indicating the foundation for the rise in excitement. Nervousness about being interviewed by police, for example, can trigger an excited response to even the most banal questions, even for an innocent person.
Furthermore, physiological events that occur because of emotional excitement are not uniform across people either in type or degree. Put differently, under pressure one person’s respiratory rate might increase by a factor of two while their cardiovascular rate remains only slightly altered or completely unchanged. Someone else might respond the opposite way. A third person may see elevations across all metrics. There is no set way any given person will respond.
Another study exhibited one reason why by examining the brain directly. By observing fMRI data of subjects told to lie, the researchers noted the areas of the brain that light up. They concluded that even at that level of analysis, far more intensive than that done in a polygraph examination, their results “do not support the use of fMRI… to detect deception in an individual.”
Mapped brain activity illustrated substantial differences among subjects telling lies. People weaving mistruths did not consistently employ the same parts of the brain to do it. The methodology for constructing a lie is simply not uniform from person to person. The obvious extrapolation from this is that if the brain function used for lying differs from person to person, surface level physiological examinations can provide no useful data toward distinguishing a lie from a truth. This is because the brain itself generates the lie, but it does not routinely use the same internal mechanisms to do it. Likewise, the brain controls physiological function. If the source of the generation of a lie cannot be definitively mapped or predicted, how can anyone credibly assert any meaning behind the activation of physiological functions?
Scientists consider the CIT to be more accurate than the CQT. It nonetheless has its own problems, in addition to the physiological ones outlined above. First, CIT is subject to false positives. Innocent examinees might learn information about a case (correct or not) through social media or rumor. Their results would, therefore, indicate that they have knowledge of a crime that has nothing to do with their involvement in it. Repeated interrogation might convince an innocent examinee that they ‘know’ something about a crime that they do not.
Second, CIT is equally subject to false negatives. An actual perpetrator may not remember details of the crime even when presented with evidence from the crime scene, forestalling any reaction from them. This is especially probable for drug-induced crimes.
Perpetrators may not exhibit any response to the type of stimuli offered in CIT examinations for other reasons. Some might simply be unmoved by seeing the tools or results of their handy work. Others might use physical countermeasures to thwart the exam. Movies sometimes show variations of this, such as when an examinee hides a thumbtack in his boot and steps on it when asked certain questions. The pain reaction manifests the same as a deceptive reaction, thereby throwing off the results.
In short, polygraphs give an appearance of credibility to a process that is entirely subjective. It is akin to holding confidence in the concept of thetans simply because Scientologists purportedly measure them with e-meters. Both e-meters and polygraphs serve no purpose other than to mislead people who believe in their randomly generated results. The problem is that polygraphs cause real-world harm.
Polygraphs are dangerous
In 2005, a study conducted by the British Psychological Society found that “polygraph examiners classified up to 47 percent of innocent subjects as guilty.” Speaking about the use of the polygraph by US federal agencies, Donald Krapohl, special assistant to the director of the Defense Academy for Credibility Assessment (DACA, formerly the Dept. of Defense Polygraph Institute) said, “We’d like the accuracy rate to be better, But it is simply the best we have.” That is ominous at best.
Steven Drizen, legal director of the Northwestern Law School’s Center on Wrongful Convictions, and an Innocence Project consultant, explained how the polygraph railroads innocent people:
The problem is that an innocent suspect volunteers to take the test and, when told that he failed, reaches the point of hopelessness where he can be easily persuaded to confess.
Researchers have shown that wrongful convictions occur more frequently when law enforcement authorities harbor preexisting beliefs about a case or subject. Eitan Elaad et al. showed that polygraph examiners tended to find deception in the readout of polygraph exams far more often when they learned that the subject confessed. On the other hand, when later information demonstrated that an examinee was innocent, examiners did not back down from their determinations that the subject was lying during the test.
In 2017, Taiwanese authorities ordered the re-examination of polygraph results conducted over 20 years by a single examiner. This came about from numerous cases that appellate courts overturned as wrongful convictions, one of which resulted in the execution of an innocent young man. One of the authors of the report on the findings of this re-examination concluded:
We found there is no standard procedure and no regulations for polygraph testing. The use of the results in criminal trials infringes on the rights of the accused and violates other human rights issues.
The Norfolk Four spent twenty years in prison for a rape and murder they did not commit because police falsified polygraph results to show that they lied in their interviews. This led one of the four to confess to the crime and implicate the other three. DNA and a confession by the actual killer finally secured their release.
Police interviewed the Green River Killer in 1984. He easily passed a polygraph, so the police released him. Afterwards, he went on to kill at least 40 more women while police looked elsewhere for a suspect. Similarly, John Arthur Ackroyd went free after raping Marlene Gabrielsen. Police let him go because Ackroyd passed a ‘lie detector’ test while Gabrielsen failed. Even though she was the victim, with evidence on her clothes, police did not bother to check the physical evidence after she failed the test. Ackroyd ended up in prison for raping and killing another woman a year later, and was a suspect in the murder of his stepdaughter.
Polygraphs are also used to exonerate people. Many of these examples make the case for why polygraphs should not be used at all. One study found that 44 pre-trial polygraph examinations supported the innocence of the defendant, but “prosecutors did not believe the validity of the results and obtained convictions from juries.” The author of this study characterized this as a fault of the prosecutors. But, in the very next section the author illustrated precisely why prosecutors do not rely on these test results. He wrote:
For 31 other exonerees, all of whose polygraphs were taken before trial, the results were inconclusive or uncertain. It could be argued that even these results should have been considered by prosecutors as pointing away from the exoneree, given the presumption of innocence, and should have given prosecutors pause before proceeding to trial or attempts to obtain guilty pleas.
This author wanted prosecutors to rely on polygraph results when it later turned out they were correct. When the results were inconclusive, however, he also wanted them to “consider” the results. This report does not comment on what prosecutors “should have done” when the polygraph results are proven wrong.
And this demonstrates the crux of the problem with polygraphs. They are eminently unreliable. Given that investigators or prosecutors have no way to know whether they are correct, what value is there in using them at all? At best, the results will be right some of the time. Otherwise, they might lead to false confessions and wrongful convictions, or put investigators on the entirely wrong path.
Courts have already accepted this reality in many jurisdictions. In 1998, the US Supreme Court majority wrote:
[T]here is simply no consensus that polygraph evidence is reliable. To this day, the scientific community remains extremely polarized about the reliability of polygraph techniques. Some studies have concluded that polygraph tests overall are accurate and reliable. Others have found that polygraph tests assess truthfulness significantly less accurately–that scientific field studies suggest the accuracy rate of the “control question technique" polygraph is “little better than could be obtained by the toss of a coin," that is, 50 percent.
The approach taken by the President in adopting Rule 707—excluding polygraph evidence in all military trials—is a rational and proportional means of advancing the legitimate interest in barring unreliable evidence. Although the degree of reliability of polygraph evidence may depend upon a variety of identifiable factors, there is simply no way to know in a particular case whether a polygraph examiner’s conclusion is accurate, because certain doubts and uncertainties plague even the best polygraph exams. [Citations omitted].
The Court was right then and still is now. Law enforcement should stop using these machines. They are unreliable, misleading, and dangerous.
***
For a story on questionable practices in the US justice system, click below.
***
I am a Certified Forensic Computer Examiner, Certified Crime Analyst, Certified Fraud Examiner, and Certified Financial Crimes Investigator with a Juris Doctor and a Master’s degree in history. I spent 10 years working in the New York State Division of Criminal Justice as Senior Analyst and Investigator.
Today I work both in the United States and Nepal, and I currently run a non-profit that uses mobile applications and other technologies to create Early Alert Systems for natural disasters for people living in remote or poor areas. In addition, I teach Tibetan history and culture, and courses on the environmental issues of the Himalayas both in Nepal and on the Tibetan plateau. For detailed analyses on law and politics involving the United States, head over to my Medium page.