Lie detectors for sex offenders?

[Big Les]

Trying desperately to leave politics out of this, can anyone see any merit in using lie detectors to assess whether paedophiles are likely to reoffend? I'm not sure I can!

 

[Darat]

No.

I heard someone in an interview on a Radio 4 programme a few weeks ago saying they knew there was no strong evidence that lie detectors were reliable at detecting lies but they still made people more truthful....

 

[Faustina]

No, lie detectors are unreliable, especially in detecting deceit in people with sociopathic tendencies. I don't think it's too much of a stretch to lump most pedophiles in that group.

Here's an interesting article.

 

[CFLarsen]

Aldrich Ames.

 

[BPSCG]

No.

I took a polygraph once when applying for a U.S. government job that required a high security clearance. Let me say up front that I have no objection to how the test was run. There was no "Gotcha!" surprise question. The tester told me beforehand what questions would be asked, and I told him the answers and we discussed any ambiguities. For example, he asked if I had ever used any illegal drugs. "Well, yeah; in college, I smoked marijuana..." "Okay," he replied, "don't worry about that; we understand that a lot of people have done that. Have you used any other illegal drugs?" I hadn't, at which point, he told me, "All right, then the question will be, 'Other than what we have discussed, have you ever used any illegal drugs?'"

We went through the entire list of questions. There were routine questions: "Is your name...?" "Were you born on ...?" Then there were "baseline" questions: "Have you ever stolen anything?" (Of course I had - who hasn't swiped a pack of gum as a kid, or taken a pen from an office?) "Have you ever broken the speed limit?" And then there were the questions designed to find out if you were a security risk: "Are you applying for this job for the purpose of betraying the United States?" ""Other than what we have discussed, have you ever committed a crime?"

There would be three run-throughs, and the only unpredictable element would be the order in which the questions were asked; it would vary from one test to the next.

During the first run-through, I was doing fine, though understandably a little nervous (who wouldn't be?), until he asked me one of the "meat" questions and I suddenly realized, "OMG, if I don't get this one, I can kiss the job good-bye!" And I felt my heart start pounding, and my face get flushed, and I knew I'd blown it.

The tester did not react, simply continuing with the test. When it was done, he told me, "Okay, you know you had a problem with that question." And we discussed what had happened; I explained that I had answered truthfully, but the sudden realization that that was a "get-it-right-or-lose-the-job" question had made me react.

We ran through the test two more times, and each time we got to that question, I reacted, though I did better the two subsequent times.

Needless to say, I didn't get the job.

I have no quarrel with the manner in which the test was run, and I have no quarrel with the decision not to hire me, even though I was utterly truthful throughout the test.

But I will never take a polygraph again, because I am certain I would react the same way. Frankly, I don't understand how anyone who understands how a polygraph works would not fail.

So, hook a sex offender up. Ask him, "Do you plan to molest little boys again when you get out of prison?" Then watch him react. Then ask yourself, "Is he reacting because he's uncomfortable with the lie, or is he reacting because he knows that he's in big trouble if he's perceived as lying?"

ETA: This should in no way be construed as a defense of sex criminals.

 

[666]

If currently available "Lie Detectors" were accurate, why not use them to replace all judicial procedures, not just sex cases? Another NO vote here.

And like BPSCG I have no love of convicted sex offenders.

 

[tim]

Lie detectors don't work. http://www.polygraph.com/

 

[Gord_in_Toronto]

Lie detectors don't work. http://www.polygraph.com/

There used to be a website with a paper that went into detail as to how a "lie detector" actually "works". I don't have the time right now to search for it again but the recommended approach, if you were asked to take a test, was to say you had read the paper paper and knew the truth of the process, they wouldn't bother testing you.

 

[fuelair]

No.

I took a polygraph once when applying for a U.S. government job that required a high security clearance. Let me say up front that I have no objection to how the test was run. There was no "Gotcha!" surprise question. The tester told me beforehand what questions would be asked, and I told him the answers and we discussed any ambiguities. For example, he asked if I had ever used any illegal drugs. "Well, yeah; in college, I smoked marijuana..." "Okay," he replied, "don't worry about that; we understand that a lot of people have done that. Have you used any other illegal drugs?" I hadn't, at which point, he told me, "All right, then the question will be, 'Other than what we have discussed, have you ever used any illegal drugs?'"

We went through the entire list of questions. There were routine questions: "Is your name...?" "Were you born on ...?" Then there were "baseline" questions: "Have you ever stolen anything?" (Of course I had - who hasn't swiped a pack of gum as a kid, or taken a pen from an office?) "Have you ever broken the speed limit?" And then there were the questions designed to find out if you were a security risk: "Are you applying for this job for the purpose of betraying the United States?" ""Other than what we have discussed, have you ever committed a crime?"

There would be three run-throughs, and the only unpredictable element would be the order in which the questions were asked; it would vary from one test to the next.

During the first run-through, I was doing fine, though understandably a little nervous (who wouldn't be?), until he asked me one of the "meat" questions and I suddenly realized, "OMG, if I don't get this one, I can kiss the job good-bye!" And I felt my heart start pounding, and my face get flushed, and I knew I'd blown it.

The tester did not react, simply continuing with the test. When it was done, he told me, "Okay, you know you had a problem with that question." And we discussed what had happened; I explained that I had answered truthfully, but the sudden realization that that was a "get-it-right-or-lose-the-job" question had made me react.

We ran through the test two more times, and each time we got to that question, I reacted, though I did better the two subsequent times.

Needless to say, I didn't get the job.

I have no quarrel with the manner in which the test was run, and I have no quarrel with the decision not to hire me, even though I was utterly truthful throughout the test.

But I will never take a polygraph again, because I am certain I would react the same way. Frankly, I don't understand how anyone who understands how a polygraph works would not fail.

So, hook a sex offender up. Ask him, "Do you plan to molest little boys again when you get out of prison?" Then watch him react. Then ask yourself, "Is he reacting because he's uncomfortable with the lie, or is he reacting because he knows that he's in big trouble if he's perceived as lying?"

ETA: This should in no way be construed as a defense of sex criminals.

I understand how it works, I passed the only three(seperate) I ever took, I lied on two of them. You have to "believe" while you are answering (mental set) and as inconspicuously as possible tense slightly on safe questions to build a tense/reactive baseline. Without going into detail, no I am not a criminal and no, I was not setting up something for any criminal purpose. Just wanted to see if I could do it based on my knowledge of how polygraphs worked .

While adding no important data, I had what was referred to as Black clearance where I was posted in the first part of my military esperience which meant I could go "behind the fence" on base and tote big flasks of biological warfare material and work in classified labs.

 

[Big Les]

Thanks for the responses. Good to know that my government really has the quality of our criminal justice system at heart, and isn't just trying to score lame points in reponse to recent bad headlines about "lost" sex offenders.

 

[Rrose Selavy]

It seems a typical ill conceived knee jerk reaction but even if they don't work, if sex offenders believed they do would it not have an effect in modifying their behaviour in some cases?

 

[Big Les]

Well, perhaps. Bit too close to the argument that X bit of woo should be welcomed because it makes people feel better, for my tastes.

 

[Avita]

Aside from the good issues raised above, there is also the question of whether sex offenders can accurately predict their own future behavior. I would bet that, like most people who are asked to predict their likelihood for socially undesirable behavior, sex offenders may sincerely answer "No, I won't offend again," but will go on to do so. And from what I understand, many sex offenders also have difficulties with impulse control.

I also think that if lie detectors actually became widely used with criminal populations, it would take no time at all before everyone in prison was swapping stories on how best to beat the machine. So I doubt that sex offenders would be afraid to lie.

 

[mist]

I believe the only actual legal use a polygraph test has is when the subject admits in it he was lying and confesses.

The polygraph test works by the tester asking you questions to which some he will expect truthfull answers, and some lied answers (he will expect you to have done marihuana, and at some time breached the speeding limit). While doing the test he will record what the difference is in your physiological reactions between truthfull and false answers. Between these "control" questions he will ask real questions like "Did you ever commit espionage" etc. and compares it to both the truthfull and false control questions.

The physiological results are somewhat within your control. For example if you would bite your tung during the control questions to which the tester expects lied answers, you can omit doing this during the real questions and it will look like you answer them truthfully.

 

[BayAreaGuy]

Great thread, Big Les! You bring up two interesting topics for debate and discussion.

I'm a Psychology professor, and I teach a short module on the Polygraph in my 101 classes. Here's what I can tell you on that subject. First, Polygraphs do not intend to, nor do they, "detect lies." They are called Polygraphs because they record several physical phenomena (as the name suggests - poly meaning more than one, and graph meaning to record or write). Most modern Polygraphs record three physical reactions: 1) respiration (breathing); 2) heart rate; 3) perspiration (microbeads of sweat on your finger tip). These have been shown in the literature to reliably change when people feel the emotions usually associated with lying (e.g., guilt, fear, panic, shame, etc.), and the machine simply records the time such changes occur, and the degree of change (deviation).

Here are the problems that plague the Polygraph as a reliable tool. First and foremost, some people simply do not consistently feel the emotions most of us do when we lie. It could be biological (as is suggested by research on Antisocial Personality Disorder - psychopaths), it could be environmental (people who have suffered long-term abuse sometimes suffer from blunted affect, for instance), and it could be purposeful (some people actually train themselves to stop feeling guilt or shame - some FBI and CIA agents still get this training).

The best, and most comprehensive, study on the reliability of the Polygraph in the courts was done in 1984. In it, researchers collected data from hundreds of U.S. court cases in which 1) the Polygraph was used to determine the guilt or innocence of the accused, and 2) the guilt or innocence of the accused was later proven with great certainty (e.g., not by circumstantial evidence).

Two frightening results were shown. First, nearly 1/4 of the guilty had been determined innocent by Polygraph testing. And, even more frightening, about 1/3 of the innocent had been found guilty by the Polygraph results.

This study caused many who had been using the Polygraph for employment or other purposes to stop, but unfortunately, it is still used by many today (as is shown in the above posts from other forum members). A 2003 study by the National Academy of Sciences was even more disturbing. You can find their results HERE. Yet another impressive study on the Polygraph which was published by the U.S. Office of Technology and Assessment is just as damning. CLICK HERE to see their results.

 

[Darat]

It seems a typical ill conceived knee jerk reaction but even if they don't work, if sex offenders believed they do would it not have an effect in modifying their behaviour in some cases?

Unusually this is not a knee-jerk response - they've been threatening to do this I think for about 2 years now.

 

[666]

It seems a typical ill conceived knee jerk reaction but even if they don't work, if sex offenders believed they do would it not have an effect in modifying their behaviour in some cases?

Maybe they'll be too busy searching the web for porn to think about searching for information about the efficacy of polygraphs.

 

[Biff Starbuck]

I have taken and passed polygraphs in the past as part of pre-employment screening. I was motivated to look into them after the second test when I was told I had failed, but yet an hour later given my written job offer. It felt like being kicked in the stomach to be falsely accused of lying for the job I so badly wanted. They never did say they thought I really was truthful, they just continued on with the hiring process. After starting work, while sitting at lunch one day my friends and I discovered that seven out of eight of us at the table had been accused of lying during the polygraph. The polygrapher apparently would just pick something he thought you might be lying about, and push you on it. If you then admitted to having been lying, you presumably were not hired.

Based upon what I have read, my biggest objection to the test is that it actually rewards lying. When the polygrapher claims that you showed a big response to lying about your name or a number, that is just for show and to convince you the test works. Most often, the control question to see how you respond to lying is about something like theft or drunken driving. They assume that if you are an adult, drink alcohol, and drive you will: A) have driven drunk B) lie about it on the polygraph. Then since the polygrapher "knows" you are lying about that answer of no, your other answers will be compared with that lie. But what if you have never driven drunk and honestly answer no? Then all of your other truthful answers match your presumed lie, and you look like you are lying on everything. But what if your body reacts strongly to the question because you really are lying about the drunken driving? Then all of your honest answers show up as a much smaller response than the lie. You pass the test with flying colors, because you only lied about the drunken driving. You are deemed MORE honest because you lied. In my opinion, someone who lies in any part of the test should be disqualified for the job. Instead, they are rewarded for being caught lying on something deemed less significant.

This is my first post, so will have to work around posting a link. Youtube has both halves of a great 60 Minutes piece about polygraph testing from 2001 entitled "Final Exam." If you search for 60 Minutes Polygraph, you should find them both.

youtube.com/watch?v=5meTECUBSAU
youtube.com/watch?v=jPXj8NqYIXk

Sadly at TAM I was not able to mention to the Mythbusters my annoyance at their episode when they used a polygraph to test if plants could detect human emotion. Before the test, they played with the polygraph and repeated the myth that polygraphs reliably detect lies.

 

[Big Les]

Great thread, Big Les! You bring up two interesting topics for debate and discussion.

I'm a Psychology professor, and I teach a short module on the Polygraph in my 101 classes. Here's what I can tell you on that subject. First, Polygraphs do not intend to, nor do they, "detect lies." They are called Polygraphs because they record several physical phenomena (as the name suggests - poly meaning more than one, and graph meaning to record or write). Most modern Polygraphs record three physical reactions: 1) respiration (breathing); 2) heart rate; 3) perspiration (microbeads of sweat on your finger tip). These have been shown in the literature to reliably change when people feel the emotions usually associated with lying (e.g., guilt, fear, panic, shame, etc.), and the machine simply records the time such changes occur, and the degree of change (deviation).

Here are the problems that plague the Polygraph as a reliable tool. First and foremost, some people simply do not consistently feel the emotions most of us do when we lie. It could be biological (as is suggested by research on Antisocial Personality Disorder - psychopaths), it could be environmental (people who have suffered long-term abuse sometimes suffer from blunted affect, for instance), and it could be purposeful (some people actually train themselves to stop feeling guilt or shame - some FBI and CIA agents still get this training).

The best, and most comprehensive, study on the reliability of the Polygraph in the courts was done in 1984. In it, researchers collected data from hundreds of U.S. court cases in which 1) the Polygraph was used to determine the guilt or innocence of the accused, and 2) the guilt or innocence of the accused was later proven with great certainty (e.g., not by circumstantial evidence).

Two frightening results were shown. First, nearly 1/4 of the guilty had been determined innocent by Polygraph testing. And, even more frightening, about 1/3 of the innocent had been found guilty by the Polygraph results.

This study caused many who had been using the Polygraph for employment or other purposes to stop, but unfortunately, it is still used by many today (as is shown in the above posts from other forum members). A 2003 study by the National Academy of Sciences was even more disturbing. You can find their results HERE. Yet another impressive study on the Polygraph which was published by the U.S. Office of Technology and Assessment is just as damning. CLICK HERE to see their results.

Brilliant stuff, thanks. And Darat is quite right that the polygraph thing has been mooted for a while now. Perhaps "kneejerk" is the wrong word...but this is still a non sequitur response to the fearful masses though. Paedophilia is today's witchcraft, at least here in Britain. We've had both convicted offenders and innocent people attacked by pitchfork-wielding morons. Strictly "kneejerk" legislation would be the handgun ban. This is more akin to the hunting ban (whatever you thought of hunting, there's little doubt it went through via weight of largely uninformed opinion on the part of the general public.

Perhaps "spasmodic and recurring kneejerk syndrome" is more accurate?!

 

[Rrose Selavy]

Well if it's been in the pipeline the reporting of it now does seem rather er..timely.. given the recent local diffculities in the Home Office.

-