How accurate are they in determining whether someone is telling the truth?
I get pretty nervous just going to the doctor for a regular checkup–I can’t imagine what I’d be feeling if I were taking a polygraph to convince everyone else of my innocence. I can imagine my heart beating faster while answering each question, worried what the machine would “say” about me. Yet increased heartbeat/BP is what the machine measures. What would my results be, assuming I were really innocent of the crime? (I know about taking a baseline reading, but again I still think my nervousness would increase while answering a question.)
Welcome to the Boards, wfq1513! Truth is, they’re not very accurate, and many times the results are not considered admissable in court. About 15 years ago, employers could force an employee to take one for any reason, but since they turned out to be so unreliable, Congress passed a law limiting their use. Generally, all they’re good for is determining your emotional reaction to a question, not if you’re telling the truth.
I disagree, I don’t really feel like looking for a source but I’m under the impression that under most circumstances they are very accurate, sure there are a few people that can fool them and they occasionally produce false results due to the fact that they are measuring truthfulness indirectly, which is why they are rarely admissable in court. But on the whole I’ve been told they produce the correct results at a substantially higher rate than they fail.
well, since they’re in the news lately, here’s what ABC news has to say about them.
I personally would hate to be subjected to one. Innocent or guilty, I’d still not do it unless forced. That fact colors my impression not only of the test and results, but of some one refusing.
You’ll note that even the fans do not claim full accuracy, only accurate to 90%, and false posatives are more likely than false negatives (that is that some one who is not being deceptive will more likely be thought deceptive than some one deceptive falsly showing to be not deceptive)
I think a good illustration of just how accurate they are can be found in today’s news.
The police in Chandra Levy’s case want a lie detector test, but they want their expert, not Condit’s, to administer it.
Peace,
mangeorge
I’ve had a buddy who took 3 of them, and based on what happened to him, I’d say their value was worthless. He passed, but they said that because of the responses on a couple of questions, there was some doubt about how truthful he was. (Which to my mind, says that even the folks giving the things don’t trust 'em!) This all revolved around a child custody case and he was told that if he passed the test, the matter would be dropped. He took it three times, and passed as I said, and then still got charged! (I’ve seen the results of the tests, BTW, and the language used was a lot of gobblety gook that seemed to serve little purpose than to cover up how inexact the damn things were.)
How do they determine that? It implies that there is another, error free way of telling the truth; if so, why administer the test? If the results are based on “control” subjects, not people under stress, I don’t see that as an accurate sample.
Very good, cp. I wondered about that myself. Ninety per-cent of what? Notice that in wring’s link the critics only estimate a 70% accuracy. A skillful interviewer (claimed accuracy, 80-90%) can beat that.
I think a polygraph, like rumor and hearsay, can be used as an investigative tool but not as evidence.
Shades of Bill Clinton, huh? Are they investigating Condit for the disappearance of Chandra Levy, or for cheating on his wife?
Peace,
mangeorge
That’s assuming the interviewer wants accuracy. Based on comments by an ex-cop in the last polygraph thread, accurate readings are secondary to the test’s indimidation value. Because cops apparently only give polygraph tests when they’re certain someone is guilty (cf., Tuckerfan’s post), accurate exculpatory answers would seem to be counterproductive.
I’ve got an old college buddy who became a DA, and a few years the subject of lie detector tests came up at a wedding reception.
His advice to me and our other friends was “It is never in your interest to submit to a polygraph test - they can only be used against you.”
His explanation was that they always do more to arouse supspicion and create doubt than anything else. They never put any issues to rest, and even if you do pass one, it can’t be used in court to your defense.
(However, years later I have identified an exception. If you’re a congressman romantically linked to a girl who has mysteriously disappeared, and you refuse to take one, your public image will take a big hit.)
There’s always the old saw (which surely everyone has heard by now) about the police trying to elicit a confession from a suspect. They put a steel collander on his head, ran a wire from it to the photocopier and surrepticiously inserted a sheet of paper with the words “he’s lying” on it. Every time he answered a question, they hit the “Copy” button. Poor guy finally broke down and told all.
The same stuff I post whenever this subject comes up:
Worked as tech support for company selling medical sensors.
Spent many hours hooked up to pneumotrodes (respiration, photoplathysmograph(pulse/blood volume), thermistor (temperature), gsr (galvanic skin response).
Lie detectors are useless at detecting lies.
The sensors only measure physical stress. A good practicioner of Tai Chi, or many meditation techniques can consciously lower their physical stress level. People who don’t fear getting caught will not exhibit a high stress level. An innocent person who is afraid will show a high stress level and create a false positive. Etc
The equipment is so sensitive that hundreds of things can throw off the readings. There are numerous tricks to screw up the data. Simply alternate between recalling a happy childhood memory and picturing sex with a favorite celebrity. This will cause your stress level to yo-yo like crazy.
Lastly, the Xerox Lie Detector story is an urban legend. Always remember http://www.snopes.com is your friend, and is second only to Unca Cecil in fighting ignorance.
I noted that even LD’s fans only ‘claim’ 90% accuracy - IOW, even those who think they’re the cat’s pajamas admit they’re not infallible. Oppontents, according to the link, add in an ‘accuracy’ rate much less. And neither defined how they determine ‘accuracy’ (conviction? other proof?) In addition, they note that the incident of false positive was much greater than false negatives. Which means on a practical basis:
taking a lie detector and ‘passing’ will not necessarily be accepted as proof, especially by those who thought you were guilty anyhow (see recent developments in the Condit affair). Thus, even ‘passing’ one will not get folks to stop focusing on you, despite that article’s statement that false negatives are rare.
Taking a lie detector and failing, and despite the lack of any other evidence, people who wish to see you in a poor light will look at this same test as being damn near infallible. This despite the article’s statement that false positives are the more common error.
Therefore IMHO, there is no value to the person being investigated in taking one. Except, of course, that your refusal to endure this flawed test with it’s lack of potentially good outcomes, will be taken as yet more evidence of your guilt. Damn things.
You’d be pretty surprised at how dumb some criminals can be. My father recently had a break-in and the stupid thief was smart enough to grab all of my stepmother’s good jewelry and left the cheap stuff behind. OTOH, she stole several semiautomatic handguns but didn’t take a single magazine, all of which were stored in the same drawers, immediately next to the appropriate pistol. She also cut the crap out of her arm on the front door when breaking out the glass and bled all over the house but didn’t bother to clean up a single drop. She did not have a valid driver’s license or other photo ID but attempted to sell the loot at several pawn shops. Definitely a dim bulb. This is more a rule than an exception. Brilliant criminals are mostly the stuff of fiction. The average crackhead burglar isn’t exactly the model of genius.
Of course crooks do dumb things, sewalk. That’s often how they get caught.
What I meant was that cops, like so many of us, like to bullshit. They simply have more fodder than most, and are not above embellishing or making stories up.
A cop once pulled me over and safety checked my old Rambler. When he found nothing wrong, he jokingly threatened to write me a ticket for driving a car that was “Felony Ugly”.
Peace,
mangeorge
I have read several articles about beating a lie detector test. One method is to place a pin inside your shoe and when your are answering each question, jam your big toe right into it. a bit extreme, but the sweat, rise in tension, and rise in blood pressure will completely throw off the test. not making you pass, but making the test results indeterminate.
I know for myself, even if i managed to be completely calm for a lie detector test, when the incriminating questions came up, i would be so nervous about whether or not the results would be accurate that i would probably fail.
I have known some liars that do it so much and so well, that they believe their own story.
