5X5 Episode 80

From SGUTranscripts
Jump to: navigation, search
5X5 Episode 80
28th October 2009

Transcript Verified Transcript Verified

5X5 79 5X5 81
Skeptical Rogues
S: Steven Novella
B: Bob Novella
J: Jay Novella
E: Evan Bernstein
M: Mike Lacelle
Download Podcast
Show Notes
Forum Topic


You're listening to the Skeptics' Guide 5x5, five minutes with five skeptics, with Steve, Jay, Rebecca, Bob and Evan.

S: This is the SGU 5X5 and tonight we're talking about polygraphs and Mike [Lacelle] is sitting in for Rebecca. Polygraph or lie detectors—although a polygraph is probably a more accurate term—is a measure of physiological responses used in interrogation in order to help determine if someone is telling the truth or being deceitful. The term "polygraph" refers to the fact that multiple graphs are taken—or measurements are taken; usually blood pressure, heart rate, respiration and sweating. So the test really is not a lie detector—it is detecting physiological responses to stress and then inferring from that that somebody is stressful because they are lying.

Now, there are various ways in which this test is done. There are different types of questions that are asked and those types of questions are used for different types of tests. The types of questions are the relevant question—a relevant question is like, "are you the one who stole the money?"—it has to do directly with the crime that's being investigated. There's various types of comparison question. A comparison question is when you try to get the person to lie, to see what their response is to lying. There are two types of comparison questions: one is a probable lie comparison, where you ask them a question that most people will lie to ("have you ever stolen money"); you can also do a directed lie comparison, where you direct the examinee to tell a lie. You say, "give us a wrong or incorrect answer to this question" and then you ask them a question. There's also an irrelevant question—an irrelevant question is something that's completely unrelated to what's going on in the investigation and is not meant to be emotion-evoking. Those are the three basic types of questions: the test question, or the relevant question, and then a control lie question, and a control non-stressful question. Then you compare the responses of the person to those three different types of questions. You want to see: when we ask them the relevant question, do they respond the same way they do when they tell a directed lie or when they answer a irrelevant question.

There's a couple of different kinds of tests that are used. One is a stimulation test, and that's basically where you're comparing the relevant question to a directed lie comparison. There's also the "guilty knowledge" test, which is interesting. There the examiner (the detective, for example) will know some bit of information about the crime that would only be known to the investigators and to the person who's guilty; the perpetrator. And then they ask a series of questions, throwing in the relevant question in the middle, and see if the person responds to that question with guilty knowledge. Again, the only reason why somebody would respond to that one question out of the cluster of questions is because they had knowledge that only the perpetrator would have. And there's also a silent answer test where you instruct the examinee not to answer the question. The theory there is that the stressful response will be greater if they are not allowed to speak; if they have to hold it in.

B: The results of the polygraph, obviously, do not show a blinking red light when somebody's lying. It takes a skilled interpreter to evaluate the results. Also, there are people who have a certain level of control, for various reasons, to the responses that the polygraph is trying to track. And there are those who simply just do not find lying a stressful experience. So beyond that, stress can be caused by more than just lying, including the stress of the test itself. A good examiner will try to account for those things. So because of these and other reasons, polygraphs are not fool-proof and they're not allowed in many courts of law in the United States, and remain to this day highly controversial. But in the right hands though, it's important to say that—in the right hands and in the right situations they can be a valuable tool for detecting deceptive behavior.

J: You know Bob, the Department of Energy was required by law to propose by April 2003 modifications to its polygraph policy, taking into account the findings of the National Academy of Sciences polygraph review. The NAS report was that polygraph screening was completely invalid as a diagnostic instrument for determining truth regarding things like terrorism, espionage, past activities of job applicants and other important issues currently so-assessed by various federal, state and local governments. The NAS found that the majority of the polygraph research was "unreliable, unscientific and biased." Amazingly, the DOE rejected the NAS' findings and proposed to retain the existing polygraph program without change.

M: As far as its admissibly in court, it various on region. If we're talking about the US it's the decision of the state or the trial judge in federal cases. In most European countries, polygraphs aren't considered reliable enough to be used in police investigations, and in most court cases involving polygraphs, they're voluntarily taken by the defendant at their expense to prove their claims. In Canada, the Supreme Court has decided to reject use of polygraph test results as evidence, but it's still used as an investigative tool. And the High Court of Australia hasn't considered its admissibility as evidence yet, but there was a case in 1982 where the judge rejected the test results on reasons including: that they were devoid of any proof or accepted scientific basis; the evidence of the polygraph operator is hearsay, which is inadmissible. So basically polygraphs are looked at pretty skeptically by most court systems.

J: Do you guys know who's considered the father of the modern polygraph? A gentleman by the name of William Marston, who under the pen name Charles Moulton, created the comic-book character Wonder Woman, and if you recall, Wonder Woman had a magic lasso that would cause anyone trapped in the lasso to tell the truth; they were compelled to tell the truth at all times. Back in the 1930s, when William Marston was coming up with his lie-detection device, it was tested out by the FBI, who actually investigated it because they wanted to see if it was a device that they would certainly be interested in using. And they felt at the time—at least a few of their special agents in charge who were researching this fellow and his device—said that the lie detector that he came up with, his polygraph, was not valid; it doesn't work, and that the claims that Marston was making—he misrepresented himself, apparently, on many many occasions. And it's kinda ironic that the FBI, which viewed the creator of the lie detector as kind of this crack-pot and a phony, and today they rely on this device to screen applicants and employees and for other uses as well.

S: But it should be pointed out that oftentimes investigators want to use the polygraph for interrogation, not because it is a lie detector, not because it tells them if the person is being deceptive or not, but as an intimidation tool, as basically a technique to get the person to crack and to confess. And that's why they want to continue to use it despite the fact that the results of it can't be even admissible as evidence in a court of law.

It is an interesting technology; it is certainly not a lie detector. It has—various studies show it's maybe somewhere between 60% and 80% accurate. A lot of it depends upon the specific techniques that are used, the skill of the examiner and, of course, the skill of resisting examination of the examinee. There are also—the door is wide open for bias. A biased examiner can pretty much get the results that they want to get. So, having a blinded examiner, of course, is critical to the reliability of the test. The question is: is a 60% accurate test of any value? Does it have any predictive value? It certainly isn't enough to be admissible in court because it's highly prejudicial. It makes it seem like if the polygraph says one thing, the fear is that the jury will just buy what the polygraph says, as if it's 100% accurate and won't put it into the proper context. But lack of 100% accuracy is not—doesn't mean it is completely ineffective and that it provides no information. It's just a really tricky tool. But it may be the kind of thing that as the technique and the technology evolves, it may get a little more accurate and may continue to have some use in interrogation and questioning. So I don't think that it's going to completely go away any time soon.

S: SGU 5x5 is a companion podcast to the Skeptics' Guide to the Universe, a weekly science podcast brought to you by the New England Skeptical Society in association with skepchick.org. For more information on this and other episodes, visit our website at www.theskepticsguide.org. Music is provided by Jake Wilson.

Navi-previous.png SGU HRes Logo sm.gif Navi-next.png