Law enforcement officers, secret agents, and counter-espionage personnel have most interesting toolboxes. Their occupations center around discovering “the truth” (or a convincing substitute) in environments where truth is scarce, and consequently they make use of methods which attempt to coerce, deceive, or scare the truth out of those who may possess it.

One of the most common interrogation tools in the history of the trade has been the lie detector. Over the years, these machines have helped put people in prison, destroy careers, and possibly even end lives. The Cold War was the heyday of these paranoia-driven truth-hunting techniques, and since the 1980s their popularity in the U.S. has declined. But they’re certainly not gone… if you ever apply for a job with the federal government, particularly a three-letter agency (FBI, CIA, etc), you’re likely to be subjected to a lie detector test, known in the industry as a “polygraph test.” But can these machines actually do what they claim?

In the recent past, polygraph tests were administered with analog polygraph machines that produced their output on paper chart recorders with colored pens; but these days they usually run as software on laptop computers with plug-in polygraph peripherals. Sensors which monitor physiological variables such as blood pressure, heart rate, respiration and skin conductivity are placed on the individual being tested (the “subject”) before questioning begins. Some units include sensors to monitor arm and leg movements as well. The data from these sensors is fed to the monitor, which records them for the polygraph operator.

Before the test is administered, the polygraph operator conducts a pre-test interview to help establish a connection with the subject, and to gain information to be used in forming “control questions.” Control questions are questions which the polygraph operator believes the subject will lie about, such as “Have you ever lied to get out of trouble?” Such questions are intended to establish a baseline reference against the responses to other questions.

The operator will then explain the polygraph machine, informing the subject that it can detect when he/she is lying, emphasizing the importance of answering truthfully. The test then begins, and includes a series of alternating question types. Some questions are the control questions; others are “irrelevant,” such as, “Are the lights on in this room?”; and the rest are questions regarding the investigation at hand. Any time the device records a physiological response greater than that exhibited with the “control” questions, the subject is assumed to have lied.

The test’s post-interview, conducted after the machine is turned off, is often the most telling portion of the process. At this point the interviewer will ask questions in such a way that the subject may believe that he/she has been “caught” by the machine, which may cause them to admit to wrongdoing.

Scientific studies of polygraph results have shown that the machines themselves provide unreliable findings. Physiological response is simply not a reliable indicator of dishonesty, but rather, it indicates increased emotions. For instance, when an important question is asked in an interview, a subject may experience anxiety at the possibility of a “false positive,” causing an emotional response. The reliability of the tests is also cast into doubt by the numerous countermeasures one can use to skew the results.

Findings imply that the success of a polygraph test is largely dependent on the subject’s belief that the technology can discover their lies. If a person is relaxed and has no anxiety of being caught, they can usually lie without detection, so long as they maintain their lies in the post-interview.

A turning point in polygraph usage came in 2001 when a number of workers at the U.S.’s major nuclear weapons laboratories began refusing the congressionally-mandated polygraph tests. A number of scientists and engineers from Sandia National Laboratories, Los Alamos National Laboratory, and Lawrence Livermore National Laboratory lost their security clearance as a result, and the U.S. Department of Energy was suddenly lacking the resources to deal with a nuclear weapons emergency.

That industry’s most outspoken opponent of polygraph screening was Dr. Alan P. Zelicoff, a senior scientist at Sandia’s Center for National Security and Arms Control. He is also a medical doctor, so his opinion carried even more weight. “Every first-year medical student knows that the four parameters measured during a polygraph⁠⁠—blood pressure, pulse, sweat production and breathing rate⁠⁠—are affected by an unaccountable myriad of emotions,” he explained, “But there is not one chapter, not one, in any medical text that associates these quantities in any way with an individual’s intent to deceive.”

Nowadays, very few courts admit polygraph results as evidence, because their accuracy is far too questionable; and laws put in place in the late 1980s prevent private employers in the U.S. from using polygraph tests in employee screenings. Yet the FBI and other federal agencies still utilize the testing widely, despite the very strong evidence that the results are unreliable, and practically useless. Feel free to draw your own conclusions there.