Hrm… my recollection was that he said he’d resign before taking a drug test. I think Reagan had proposed drug testing for all, or a large number of, government employees.
From the depths of my cobwebby memory banks, I seem to recall an alternative to the polygraph called the Voice Stress Analyzer, or VSA, which was supposed to be far more accurate. Was that hype that did not withstand closer examination, or what? If they work, why don’t we hear about them any more?
Read this excerpt from the Skeptic’s Dictionary.
It was definitely hype, as the the VSA is no more (or less) accurate than the polygraph. Both devices are a fraud.
I was asked in a polygraph whether I had smoked pot in the last 5 years. Well I hadn’t smoked it for close to 10 years and even then only in experimentation. But the whole subject made me extremely uncomfortable to talk about and the examiner noted I didn’t do well on the question. He didn’t say I was lying but he inferred I had a problem and I was glad I accepted another job and didn’t have to face the interviewer who sent me to get the polygraph again because I don’t know if he would have called me a liar or not.
I will never take a polygraph again. They may detect emotional variations, but I don’t think they can specifically tell when someone is lying.
No, it was lie detector. Apparently there was some kind of a flap about leaks to the press and Reagan said something about subjecting members of his staff to lie detector tests to find out who was leaking.
I try to assume as little as possible. I have no intention of changing that policy simply because you ask me to. You say that everything I have ever read about polygraph tests is false, and what do you have to defend your position? Thorough scientific evidence? No, of course not! I’m supposed to just take your word for it! If we’re going to be asking each other to make assumptions, here’s one I’d like you to make: please assume that I am not a complete idiot who will believe anything you say.
Can you find anything in those links which disputes my claim that 1 - beta > alpha ? Yes, scientists disagree about how effective polygraph tests are, but that doesn’t mean that they disagree about whether they are effective.
What does the fact that you started the thread have to do with it? What am I supposed to say? “Gee, if Hardcore started a thread on this, he must know everything there is to know about it”?
What conflicting evidence?
Seeing as how you haven’t presented any evidence, I guess it is accurate to say that I am unaware of it.
It’s always nice to have a discussion with someone who can interpret disagreement as “insolence” :rolleyes:.
hardcore wrote:
DO NOT QUESTION THE WISDOM OF MAGIC 8-BALL!
tracer :D:D I’ll try not to mock Happy Fun Ball either.
The Ryan
I have never made such an assumption, and likely never will. However, I will assume that you haven’t bothered to research the links I mentioned. So I will go through the tedium of pasting portions of it on this page to increase the chances that you might read it. All bolding will be mine.
Since this sounds greek to me, I will assume you are referencing the subject of this thread – the scientific validity of polygraphs as lie-detectors. Let’s start with a few excerpts from the first link, The Skeptic’s Dictionary:
I realize you might dismiss the information from The Skeptic’s Dictionary because it is just so, well… skeptical. So let’s try the next link provided, from our own government’s Office of Technology Assessment, entitled Scientific Validity of Polygraph Testing: A Research Review and Evaluation
Just in case you think the government is in on the conspiracy, I’ll go to the next link offered, Psychologists debate merits of the polygraph:
To be thorough, I also include this from the aforementioned Statement of Steven Aftergood:
Now all of the above was included in the links JonF provided and I mentioned to you. Yet you replied to me with this:
Hmmm… are you now aware of it? Or do you simply proclaim everyone cited above to be “ignorant at best”?
Always one to try and provide more than necessary, I present further evidence of the polygraph’s fallibility. To determine the scientific community consensus, see The Validity of the Lie Detector: Two Surveys of Scientific Opinion
Since the FBI likes to use the damn thing, I thought this statement from the Head of the FBI Polygraph Unit, Special Agent James Murphy would be appropriate.
A recent article in Salon (admittedly not a scientific journal) had these tidbits from The truth about the polygraph
Please tell me you do remember Edward Teller.
For additional statements from scientists, I now turn to a 1983 Federal study of polygraph accuracy, where the author of the study, Brandeis University psychologist Leonard Saxe, PhD, stated the following about the polygraph
Likewise, in the Opening Statement of Dr. Drew C. Richardson, FBI Laboratory, before the United States Senate, he had this to say about polygraphs:
An article from Al Seckel at CalTech entitled Pinocchio Science: The Truth About The Polygraph states
While John J. Furedy, Ph.D., Professor of Psychology at the University of Toronto simply states:
Now, you might be able to uncover something by Raskin or Honts to support your viewpoint. Yet in the midst of this veritable mountain of opposing data, you assert
I’m sorry, but in light of the evidence above, I have to conclude that it is you who are “ignorant at best” of the current scientific status of polygraphs.
Hardcore, I just don’t get your point here. Does this thing work or not? Try to get off the fence in your next post.
Thanks. hardcore; I considered going through the process that you went through, but decided I didn’t have time. Looks as if you did it more thoroughtly than I would have.
So as an employer, you should ask people to take polygraph tests. If you are looking for intelligent well educated people, hire the ones who refuse. If you are looking for sheep, hire the ones who take it. 
picmr
Just a late addition. This last week “Good Morning Amerika” reported the 95% accuracy quote as an absolute fact. So now 15 million or so additional people will believe it.
It is standard statistical notation.
Now, let’s see what “evidence” you have:
Seeing as how polygraph tests are not concerned at all with causation, but only with correlation, that makes no sense.
So? Do they have evidence for this statement?
Sounds better than random to me.
No, I still have yet to see any evidence that the polygraph test does not work. All I see is a bunch of quotes from people that haven’t seen any evidence that it works. Ansence of evidence is not evidence of absence.
I agree with this policy.
The fact that it leads to false positives hardly means that it doesn’t work.
Just to make my position clear: I don’t think that polygraphs are perfect or even good instruments for determining whether someone is telling the truth. But given a choice between a Magic Eight Ball test and a polygraph test, I’d definitely go for the polygraph test.
Hey, Ryan, how’d you get your post to change to a Times Roman font near the end there?
Actaully, you have seen a bunch of quotes from people who have looked very hard for evidence that it works, and found none; but they have found evidence that it does not work.
Got any citations for scientific evidence (statistical evidence is acceptable) that it does work?
Tracer:
Doesn’t look like a different font to me.
JonF
“Looked very hard”? What are you talking about? Just what have these scientists done to look for evidence? Have they performed experiments? I haven’t seen any cited.
No, I don’t. Have I started a thread in GD proclaiming polygraphs to be effective? No, I haven’t.
No, you have not started a thread. In an existing thread you have stated:
“It is well established that there is a positive correlation between a polygraph machine saying that somone is lying and that person actually lying. The statement “These devices likely have the same effectiveness in determining trustworthiness” is ignorance at best.”
“scientists disagree about how effective polygraph tests are, but that doesn’t mean that they disagree about whether they are effective”
You have made claims without any support for those claims. Is there a difference between starting a thread with that assertion or making that assertion somewhere inside a thread?
I am talking about the references which are available to you just by clicking on the links in previous messages. Some of these are reviews of experiments, some are reviews of the literature of experiments. Many of them include references to reports of the reviewed experiments. The only reason that you haven’t seen any experiments cited is that you have refused to make the effort of clicking your mouse and reading. You want to know just what these scientists have done to lok for evidence? Read what they’ve written!
You have also refused to provide any references to support your assertion. You made a claim; do you have any support? It’s your claim, it’s your responsibility to support it, especially when evidence has been provided that contradicts your claim.
Yes. I would not start a thread dedicated to an assertion unless I had evidence at hand for that assertion.
I don’t have several hours to spend wafding through those websites. You want to see evidence that polygraph tests are effective? Go to a library. How is that any better?
It arose from a claim that hardcore made, a claim that has not been supported.
You evidently have a very strange definition of “provided”.
Actually, the evidence that hardcore presents shows there isn’t even much correlation.
Harcore cites at least two studies:
Just for reference, calling a statement a lie if a coin flips tails will give you about a 50% success rate.
Actually it does. Its results would perhaps be acceptable even if it gave a fairly high percentage of false negatives. But the presence of false positives (indicating that people are lying when they’re telling the truth) is precisely the reason why civil libertarians denounce its use. Given the high (43%) rate of false positives, a truthful person would be a fool to subject himself to the test.
A common debating technique is to offer an assertion to criticism. Generally, one should have one’s ducks in a row from day one. Hardcore might be guilty of a little sandbagging. However, the links he provided were on point. I can’t see as how anyone would be able to fairly accuse him of asking you to “go to the library.”
When someome makes an assertion, they have the burden of proof. Hardcore seems to have satisfied that burden. It is then up to anyone who wishes to contradict his proof to offer evidence to the contrary.
I’ve never had a reason to consider the effectiveness of polygraph tests. Before this thread I had given it almost no thought whatsoever. I admit a small bit of bias. The concept of a “lie-detector” seems a bit far-fetched. But I have seen stranger things proved irrefutably true, so I try to keep an open mind. Based, however, the evidence presented in this thread has begun to persuade me that lie-detectors have little more validity than the Magic 8 Ball.
Do you know what the word “correlation” means? Have you seen anything that looks like a discussion of correlation?
Barely better is still better.
So the test for HIV doesn’t work? Drug tests doesn’t work? The justice system doesn’t work? If that’s your defintion of “work”, then no, polygraph tests don’t work. Can you name a test that does work according to your definition?
I never said that the test was acceptable (except in extreme situations), I just said that it was better than a random test. That is, the rate of false positives is less than that true positives. All of the statistics that you quoted gave only one number, but both numbers are important. Tell me something. If you had a choice between a polygragh test and someone just flipping a coin, wouldn’t you prefer the polygraph? Hardcore said that they’re completely equal. That’s what I’m disagreeing with.
Do the rest of you have a different browser or something? Because I have yet to see any evidence.