How woo-woo are lie detector tests?

[ca3799]

I'm goofing off today and watching Judge Larry Joe. He gets to use lie detectors on his show.

I've taken three lie detector tests (pre employment). I believe I passed all the tests. While the examiner won't tell you the results, I was able to continue with the pre-employment screenings and did get two out of three jobs that I screened for.

Prior to the last exam I took (and it's been many years now), I asked the examiner "How accurate is this test?" His only response was a cool "Very."

To take the test, you have to answer a questionnaire. I think there are two reasons for this- one, to help the examiner frame the questions and two, to give you plenty of time to get nervous about taking the test.

For example, one question asked if I had ever stolen anything, and then went on to ask specifics. Well, I hate to admit it, but I have stolen at least one thing in my 45+ years (nothing lately, tho, honest!). To cover for this problem, I actually created another lie and stated that I had stolen a head of lettuce from a grocery store on a dare and as a teenager.

During the test I tried my best to remain calm and regulate my respirations and heartrate, especially following a question. They ask you to briefly pause after a question before answering.

So, I did lie during the test and manipulate the test to the best of my ability.

After the test, the examiner said "What are you, God or something?" I didn't know if he was teasing me knowing I had failed the test or if he was telling me I passed. I said "No sir, (Texas girl), I'm just from a good family" (another lie, hee hee).

I did not get that particular job, but I am pretty sure it was not LD test related. I was told came in third for one open position.

So, I know I was not truthful during the test and I believe I was passed.

I'm not asking if LD tests work because I obviously feel they don't- at least on me. Do you think they work? Do you think they work on susceptible people? Do you think they are a colossal waste of time and resources?

 

[Dragon]

Those of us in Europe are generally bemused by the reliance of agencies in the USA on this sort of test.

Have you had a look at http://antipolygraph.org/?

 

[davefoc]

The myth about the efficacy of lie detectors seems to be just one more example of the fact that in human cultures once a lie becomes told often enough and with sufficient sincerity, it becomes so thoroughly rooted that the culture treats the myth in the same way that it treats actual facts.

These kind of lies do sometimes become dislodged, (phrenology is a common example) but this one seems annoyingly persistent right now.

 

[pupdog]

They are most "accurate" when the operator can convince (by intimidation) the testee that they really work.

 

[mgdwcb]

One trick to beat the detector could be to turn the questioner's questions into your own question and to answer that quite honestly. I'm sure that this sort of thing is taught to agents and soldiers and that `psychological performers' like Darren Brown here in the UK could do it, if he hasn't done it already. With a bit of practice most people could probably beat the detector too.

 

[specious_reasons]

George Maschke posts to this forum every once in a while. I have to say that I'm more educated about polygraphs due his site.

Polygraphs in criminal investigations have some use, based on FUD and intimidation. Otherwise, it's yet another pseudo-science.

 

[ca3799]

Those of us in Europe are generally bemused by the reliance of agencies in the USA on this sort of test.

Have you had a look at http://antipolygraph.org/?

That is a very interesting site, and jibes with my initial impressions quite a bit. Thanks for the link.

 

[goldie]

There is no scientific evidence that lie detectors work. They only work because the testee believes they work. In addition, no one knows how many people have either not gotten jobs or lost jobs based on faulty polygraph findings. As far as the US security agencies use of these tests, it is scary that they are basing the country's security on such a test. There have been some spies in the recent past who worked for the FBI and the Defense Department who passed these tests with flying colors. Because the security agencies believed in these tests, these individuals were allowed to keep on spying until they were eventually caught (not by lie detector tests BTW).

 

[Marian]

My husband had to take a polygraph when applying for a transfer to DoJ.

It's pretty commonplace in law enforcement, both state and federal.

As most people probably know, polygraph results are not allowed as evidence in criminal court cases. The jist of it is that it's more 'art' than science. In other words, results can not always be duplicated. Which makes it...well...not very useful.

Various police agencies DO make use of polygraphs in 'eliminating' suspects, or in an attempt to get direction on a case. However, the results can also be harmful to a case. There are numerous situations where true, the police focused in on the correct suspect due to polygraph results. There are also cases where police stopped investigating a suspect (usually in cases without much evidence or large suspect pools) because of false polygraph results.

Unfortunately I think the general public lends polygraphs much more believability than they should have. They're unreliable, they're "beatable", and the results are often vague. It would certainly seem to me to be a much better idea to actually take another look at the 'science' behind it, and either junk it (which I think would end up being the case), or attempt to revamp it so that you end up with something that works better.

Or even keep it, but with the better understanding that they're not highly accurate, and may or may not be beneficial to a case (when used as an investigative tool). Certainly making calls based primarily on the results...is a bad idea. :\

 

[davefoc]

Marian, what you said is fairly consistent of what I thought the general situation with regard to the useage of polygraphs by law enforcement would be.

The scenario that I have assumed most of the polygraph situations are like is that the examiner has a fairly good idea of what the truth is based on the evidence collected so far and then uses the polygraph as a tool to pressure the suspect into admissions or at least into revealing damaging information. The examiner then either finesses the polygraph results or just lies about them to further pressure the suspect.

None of this is too bad until the police get fooled by polygraph examiner into using his information as another datapoint in their investigation. I think this is the problem you are talking about. And there have been famous cases where the federal agencies that use polygraphs do exactly this and make enourmous mistakes as a result.

I don't have a feel for is the statistics associated with this. How often is a polygraph useful in an investigation, how often does it obscure reality to the point that an investigation is derailed. I don't know what studies have been done with respect to this but it seems like a very difficult thing to study because of the confounding fact that the examiner has so much data to make truth/non-truth judgments outside the data obtained from the polygraph.

It seems pretty easy to test the efficacy of a lie detector in controlled test situations. I think these tests indicate unreliability that suggest polygraphs are not useful, which is one reason that polygraph proponents make up all sorts of excuses as to why they aren't relevant.

 

[Michael Redman]

I've taken three lie detector tests (pre employment).

Who administered these tests? In the US, Federal law prohibits the use of pre-employment polygraphs by most employers.

Of course, that does't stop them from using it themselves in security related areas. The fact that they don't see the self-evident foolishness of this double standard frightening.

 

[RapmasterT]

This is actually a subject I can lend some factual insight on.

I am a former US Army CID agent and was "fortunate" enough to be assigned to an office with a resident polygraph examiner. He and I became very good friends and I obtained a considerable exposure to the polygraph process through this friendship. I've personally been polygraphed literally dozens of times.

The common mantra of the trained examiner is "you can fool the examiner, but you can't fool the machine". Which is basically a copout, stating that you WILL have a physiological reaction when lying, but it's up to the examiner to interpret it correctly.

The part where most people get confused is the belief that there is a pass/fail or guilty/innocent shortcut through the polygraph. Hollywood is partly to blame, unethical examiners also, and many find the myth to be very useful in practical applications.

The reality is the polygraph is simply one of many interview and interrogation tools, just like telling a subject that his accomplice has already confessed, or that his fingerprints were found at a crime scene, or the ever popular "good cop, bad cop".

90% of the tests I sat through, were simple calibration scenarios. Given a number between 1 and 10, I was told to say "no" to every question of "is this your number" as they were counted off. I knew when I was going to lie, the examiner knew, and watching that number come closer and closer, it's almost impossible to NOT react when the time comes, even if you've done it a hundred times before. There's usually a HUGE spike on the chart in bright, red, accusing ink.

Every subject placed under examination was given this same "calibration" to start off. Ostentiably to set the machine up, but in reality it's a controlled demonstration of "see, you can't even lie about a simple number without us knowing it". This convinces the subject that hte polygraph will catch him lying, and hopefully triggers an extreme reaction when the nasty questions come along.

What follows is not an examination, but an interrogation where the subject has already been convinced you'll know when he's lying. The real meat doesn't come from the polygraph though, it comes from the examiners questioning and breaking down the subjects false stories. I don't know how civilian agencies choose their examiners, but the Department of Defense makes it extrordinarily difficult to become an examiner. The candidate must already be established as head and shoulders above the field as an interrogator, long before the polygraph is ever given to him as a tool.

Polygraph has legitimate criminal investigations use, just like any other interrogation technique. The overuse, misuse, and belief that it's a "lie detector" is what's likely to get it taken out of responsible hands as well as the quacks.

 

[Cynical]

Lie detector tests are very woo-woo. They don't really detect lies, they just detect physical reactions to your feelings. You can pass them if you're cool enough.

 

[ca3799]

Re: Re: How woo-woo are lie detector tests?

Who administered these tests? In the US, Federal law prohibits the use of pre-employment polygraphs by most employers.

Of course, that does't stop them from using it themselves in security related areas. The fact that they don't see the self-evident foolishness of this double standard frightening.

Two were for jewelry sales and one was for a police department dispatcher position.

 

[Marian]

Polygraph has legitimate criminal investigations use, just like any other interrogation technique.

Just an interesting aside, and one that surprised me when I heard it, according to Project Innocence (which uses DNA testing to review cases of convicted criminals) 1 in 4 convictions in which the convicted person is proven innocent (due to DNA evidence that can now be tested) has confessed to the crime.

My point being that many "interrogation techniques" are also suspect, given that false confessions take place. I know many people in law enforcement, and personally based on my experiences do not believe that most of these 'false confessions' take place due to any form of abuse, (in other words it's not beaten out of suspects, nor does it occur in 20 hour questioning sessions where the person cannot use a restroom or some nonsense), but rather because of interview techniques which can illicit a false confession. (Such as the police being allowed to lie to suspects during questioning, etc.)

I think it's obvious we can't throw the baby out with the bathwater...interviewing is a major part of criminal investigations. However given the statistic that of people cleared through Project Innocence with DNA testing that 1 in 4 have verbally confessed to the crime (many of whom also wrote out and signed confessions) then how many more people would fail a lie detector test? Remember these are people who are verbally confessing to tremendously serious crimes. I think most people believe they would never confess to a rape or murder they didn't commit, especially given the atrocious nature of such crimes.

I think if 1 in 4 cleared through DNA through Project Innocence has verbally confessed to crimes they didn't commit, then it would seem likely to me (though this is speculation, and IMO would certainly need testing) that you'd have much higher ratio of people failing polygraph testing. That high of a failure rate in confessions beggers the question of how much credence even a confession should be given in court. Obviously they're still going to be given high consideration...but it's demonstratitive I think of the need in the pursuit of justice to examine carefully the entire picture, and not merely what's expedient.

And I wonder if hard data was available on how well (or how poorly) polygraphs worked, how much various law enforcement officers would rely on the results. I think having that information would be essential for them in determining how much weight to give to polygraph results, especially in making determinations on how to proceed in a case.