Polygraph Testing - Science or BS?

[FFed]

When I took one the examiner did the old card trick where I pick a card out of the marked deck of 10 cards and she can determine which one it is by asking me. It was so pathetic and cheezy that I almost burst out laughing. Controlling my laughter and acting like it was real was harder then the actual test.

 

[marting]

A much bigger part of the reason for the continued use of the device is the inability of polygraph opponents to realize that polygraphs are substantially better than chance. Polygraph opposition seems to turn into a classic politically motivated false-dichotomy -- if polygraphs aren't error-free "magic bullets," then they're no better than chance.

The simple fact is that polygraphs are, in fact, better than chance at detecting deception, especially as part of the investigation of a specific incident. From a 1983 US government report (republished by the Federation of American Scientists) :

Emphasis mine. "Better than chance," but with significant error rates. (And technology has only improved since then,.... but the error rates remain significant.)

The report (therefore) doesn't recommend using polygraphs for screening, a recommendation with which I concur, for reasons that should be obvious from the OP. But let's not throw the baby out with the bathwater here. And let's not tell easily refutable lies.

It doesn't seem the improvements in technology since 1983 have helped much. The National Academy published this extensive review in 2003:

http://www.nap.edu/openbook.php?isbn=0309084369

They also concluded that specific incident tests (such as for a known crime) work at levels much better than chance but well short of 100%. However, they found screening tests much less reliable than that and far from the claims of the polygraph community.

To a great extent polygraphs work best with a bit of examiner deception (such as what a control question is really for). Knowledge and a good, deep understanding of the details of the polygraph process could reduce it's utility. It's an open question as to whether an innocent person is better off researching it. Personally, I like to understand stuff but then I don't have or expect to ever have a job requiring a polygraph.

It is a fascinating field, more art than science, and populated by some interesting characters (google polygraphing plants). The technique with the most science behind it is known as the "Concealed information test" which is well suited for forensic work, useless for screening, but almost never used in the United States.

 

[ImaginalDisc]

No, they are very good at detecting those physical changes which makes them able to detect lying at a rate better than chance.



(And I promised myself I wouldn't get involved after the last polygraph thread...)

I'm surprised I need to inform you about the existence of a false positive, and how a test which merely detects physiological stress can be used to erroneously label an angry, nervous, or tense person as a liar.

 

[Lanzy]

So what can you do about it? I have recently applied for a job that will definitely require I take a polygraph. I want the job and will undoubtedly take the exam if I get that far. Can you prepare for this?

 

[JoeyDonuts]

So what can you do about it? I have recently applied for a job that will definitely require I take a polygraph. I want the job and will undoubtedly take the exam if I get that far. Can you prepare for this?

I hate to tell you this, but there's nothing you can do to really prepare for it. If the examiner gets an inkling that you are actively trying to control your body's responses, he'll call foul and the test will be declared null and void.

Unfortunately, the deck is stacked against you here because you're taking a polygraph administered in a SCREENING capacity.

Let's say a murder has been committed and a suspect is being given a polygraph. If he is asked questions relating to SPECIFIC information about an actual INCIDENT that had taken place, then the polygraph could conceivably give a better-than-chance indication over whether the person had anxiety or stress about any of those points.

Unfortunately, polygraphs seem to be set up to ensnare the anxious innocent and the guilty with a conscience.

One need look no farther than Aldridge Ames to see how effective they are in a screening capacity.

If YOU must subject yourself to this test in order to secure this job, DO NOT look at anything on the web that would give you advice for how to defeat it. You do not want anything extraneous running around your brain during the test. Get a good night's rest, be well-fed and rested and try to be at ease with both the process and the examiner. (This was one of my problems, as the test administrator was a dead ringer for Montgomery Burns and gave me a serious case of the heebies.)

Most of all, if they say you have failed it (which I hope for your sake doesn't happen) DO NOT ADMIT TO ANYTHING. Do not fill out a statement, do not carry on any more conversation with the administrator other than "I would like to speak to an attorney." Your situation may be different, and God I hope so. But if you get even the slightest feeling that something isn't right, TERMINATE THE SESSION IMMEDIATELY.

There are more important things than securing a job that requires a polygraph. That said, I hope you get through it. It's nerve-racking to say the least. I should mention that I DID pass one with no problems early on in my career, before I had "access to anything."

Good luck, and watch your six - because nobody else is going to do it for you.

 

[I Ratant]

So what can you do about it? I have recently applied for a job that will definitely require I take a polygraph. I want the job and will undoubtedly take the exam if I get that far. Can you prepare for this?

.
I've had four of them.
You have to remain calm, and "bond" with the examiner.
Do NOT try any of the silly tricks to "fool" the machine. These merely raise suspicion, and can result in a retest.
All of the questions are discussed before the formal test, there should be no reason to be alarmed at any of them once the test is run.
I know people who failed by trying to trick it, by getting upset at the questions, and just being too scared of the thing to even take it.
There is no proven correlation between any of the physiological responses of the machine and truthiness.

 

[George Maschke]

I hate to tell you this, but there's nothing you can do to really prepare for it. If the examiner gets an inkling that you are actively trying to control your body's responses, he'll call foul and the test will be declared null and void.

...

If YOU must subject yourself to this test in order to secure this job, DO NOT look at anything on the web that would give you advice for how to defeat it. You do not want anything extraneous running around your brain during the test. Get a good night's rest, be well-fed and rested and try to be at ease with both the process and the examiner.

I must respectfully disagree with you here. Polygraph techniques necessarily involve the polygrapher lying to and otherwise deceiving the person being "tested," and foreknowledge of what to expect can most certainly help one through a polygraph interrogation. Knowledge is preferable to ignorance, and it would be foolhardy to remain willfully ignorant.

Detailed information about the polygraph procedure and interrogation approaches used in polygraph screening (including the thoroughly discredited Relevant/Irrelevant technique used by the NSA) is provided in Chapter 3 of The Lie Behind the Lie Detector. Chapter 4 discusses techniques that can be used to reduce the likelihood of a false positive outcome:

http://antipolygraph.org/lie-behind-the-lie-detector.pdf

The late Dr. David T. Lykken, who discusses polygraph countermeasures in Chapter 18 of his seminal treatise on polygraphy A Tremor in the Blood: Uses and Abuses of the Lie Detector (2nd ed., 1998), concludes "...if I were somehow forced to take a polygraph test in relation to some important matter, I would certainly use these proven countermeasures rather than rely on the truth and my innocence as safeguards..."

 

[Bikewer]

Some years ago, 60 Minutes set up a little scam. They took over a camera shop temporarily, and set up several staffers as "employees". Then, they called three different polygraph firms to investigate "employee thefts".
Each polygraph operator was told something along the line, "we're not really sure, but we THINK it might be employee X."
Sure enough, each polygraph operator found employee X guilty of "deception".

It's an old adage of police work that "it's all in the operator". Having watched this process on occasion, I can testify that good polygraph operators are simply those individuals who would make very good interrogators without the machine. They are sensitive to mannerisms, tics, nervousness, etc. Poker players call these things "tells".

Interesting that though polygraph results are universally barred from criminal courts, they are routinely used in police hiring.

Lately, we have been using the old (and long discredited) "voice stress analysis" technique. I actually heard one of our investigators tell a suspect, "it's infallible".

Intimidation, just as CFLarsen says.

 

[Lanzy]

Thanks everyone.

 

[drkitten]

I'm surprised I need to inform you about the existence of a false positive,

I'm surprised that I need to inform you about how the existence of a false positive doesn't render the entire system "little better than chance."

One swallow does not a summer make.

 

[Arthur Denton]

It is a nice thread this one, new woo for me, yes, this thing about the Polygraph test. I thought that the polygraph not being accepted into courts was something weird but this clears out a lot of things for me.

Well, there is a particular program in Brazilian television that uses the polygraph for tests on couples. I mean, usually the husband is in the polygraph and he's asked about being loyal, having loved a girl, etc, etc... just for the show to keep interesting to the spectators. It amuses me that at least some of those cases seem to be real. I thought all of them would be blatant frauds for the looking, opened, I mean, easy to detect forgery. But it amused me anyway sometimes, it's like those faith healing programs that plague TV these days, so sometimes I watch pieces of it. A twelve minutes polygraph test, with questions like the following, determine if the person is guilty or not of betraying his wife (husband):

1. Have you any secrets from your wife?
2. Have you ever felt desire for another woman outside your marriage?
3. Do you think your wife is as attractive as she was when you married?

I mean, even if you love and respect your wive, secrets will happen. You work, you talk to other people, so you have secrets. And look, you do the test with the person who doubt about you in front of you, so that ought make you nervous in the process, thus reducing the chances of a decent exam. Men and women can easily feel attracted by someone, and they feel guilt for it. The question is completely biased, I mean, who never looked at the side and thought someone is attractive? And as for the last question, I mean, after 15 years of marriage, of course you and your wife are not those hot cakes you once were, life does that with people, it's natural so it is a natural "no" question, and would you say that to someone you love? No, you would lie, as it is a white lie for the good of the relation.

And now there's a prize-show (imported from the US I think) where people face the polygraph on a quest for a money prize. It's amusing to a certain extent too, but many questions seem just biased to me, I mean, they are meant to confuse, and since the polygraph is a truth/lie thing, questions should be straight questions, not biased ones.

Nice thread

 

[George Maschke]

Some years ago, 60 Minutes set up a little scam. They took over a camera shop temporarily, and set up several staffers as "employees". Then, they called three different polygraph firms to investigate "employee thefts".
Each polygraph operator was told something along the line, "we're not really sure, but we THINK it might be employee X."
Sure enough, each polygraph operator found employee X guilty of "deception"....

That classic 60 Minutes report may be viewed here:

http://antipolygraph.org/blog/?p=110

 

[JoeyDonuts]

I must respectfully disagree with you here. Polygraph techniques necessarily involve the polygrapher lying to and otherwise deceiving the person being "tested," and foreknowledge of what to expect can most certainly help one through a polygraph interrogation. Knowledge is preferable to ignorance, and it would be foolhardy to remain willfully ignorant.

I respectfully aknowledge your disagreement. I can only say in my defense that I was thinking of my future career with the Navy, and unfortunately for me, that led right through a polygraph. However, had I known the hell I was going to be put through for months and months on end, I probably would have done a tad bit more homework.

Interestingly enough, in the NCIS prebriefing for the polygraph, they asked a group of us if we had done any research on the polygraph and specifically mentioned your website in a not-too-flattering light. Without mentioning you by name, they cast you as a disgraced and disgruntled person who has a bone to pick with the authorities.

I wish I would have known that this was coming from an organization that will most certainly lie to people in order to get confessions.

Ya know, NCIS headquarters looks a lot different on the TV series than it does in Norfolk. It's a damned trailer farm tucked away in a corner of Naval Station Norfolk. Also, evidence tape is really hard to get off of computers...

 

[ImaginalDisc]

I'm surprised that I need to inform you about how the existence of a false positive doesn't render the entire system "little better than chance."

One swallow does not a summer make.

Not a false positive, false positives from numerous different sources. Can you tell how often a polygraphy subject is nervous, tense, stressed, constipated, excited, upset, or in one or more of any number of conditions besides being deceptive?

 

[drkitten]

Not a false positive, false positives from numerous different sources. Can you tell how often a polygraphy subject is nervous, tense, stressed, constipated, excited, upset, or in one or more of any number of conditions besides being deceptive?

Why, yes, yes, I can. In fact, I not only can, but have, in this very thread. Complete with citation...... The cited false positive rate is about one in seven.

 

[borealys]

o correct guilty detections ranged from 70.6 to 98.6 percent and averaged 86.3 percent;
o correct innocent detections ranged from 12.5 to 94.1 percent and averaged 76 percent;
o false positive rate (innocent persons found deceptive) ranged from O to 75 percent and averaged 19.1 percent; and
o false negative rate (guilty persons found nondeceptive) ranged from O to 29.4 percent and averaged 10.2 percent.
* Fourteen individual analog studies:
o correct guilty detections ranged from 35.4 to 100 percent and averaged 63.7 percent;
o correct innocent detections ranged from 32 to 91 percent and averaged 57.9 percent;
o false positives ranged from 2 to 50.7 percent and averaged 14.1 percent; and
o false negatives ranged from O to 28.7 percent and averaged 10.4 percent.

To me, the variability in those numbers is unsettling. One of the key factors used to determine whether a test is reliable is whether or not its sensitivity and specificity are comparable across times and testers. I haven't read any of the studies summarized, and it's possible that some were better constructed and carried out than others, or that they used differently-trained administrators or different subject groups. If that's the case, those numbers don't tell us much, but with only those numbers to go by, it doesn't look like polygraph testing is at all reliable, even if, over a large group of subjects, it does come out better than chance.

If it's not reliable, it's not interpretable in any one individual case.

A diagnostic test that is not reliable is not valid. Period.

 

[jdp]

I also recently had the opportunity to apply for a position that would have required a poly. I ultimately passed on this go around because I didn't feel like I had enough time to fully investigate the whole situation, this is something I wanted to know all about before I went in.

My initial thoughts are, and have always been, that polys are BS. And that it's mostly theater set up to 'get to' the one being tested. Basically, you would be at the mercy of the tester and their mood on the given day or of their opinion of you. I was afraid of getting into a similar situation as JoeyDonuts. I am courious if there is any substance behind the claims.

ETA: I must agree with borealys the numbers posted don't really instill a lot of confidence in me that a poly is anything more than a game. Also didn't Robert Hanssen also pass several polys too?

 

[borealys]

I don't think that the occurrence of some false positives and some false negatives is reason to abandon a test. No test is perfect, and I would be more suspicious of claims that it's impossible to beat it than that it's difficult to beat it.

What's potentially damning about those polygraph numbers is that the frequency of false positives, in particular, can't be determined with any reasonable accuracy. In any given case, the likelihood of being falsely condemned as "deceptive" based on the test could be as high as 75% or as low as 2% -- which basically means that we do not know how likely a false positive is.

Given that a false positive could condemn someone to prison or cost them their career, as in Joey's case, that's disturbing.

 

[drkitten]

To me, the variability in those numbers is unsettling.

As well it should be -- although it's not clear whether the testing situations were similar enough that they should be directly comparable.

But it's one thing to say a test is not reliable. It's another to say its "only slightly better than chance" or "no better than chance."

Even unreliable tests can be useful (as the government reports have pointed out); the unreliability of eyewitness testimony has ALSO been done to death, but no cop would think of suggesting not interviewing eyewitnesses as a necessary step in incident investigation. The difference --- and this, again, is in keeping with the recommendations of the various reports --- is that cops don't tend to go to court with nothing but the testimony of eyewitnesses to support them. (When they do, they tend to lose.) Eyewitness testimony tends to tell them where to look for other sources of evidence.

 

[George Maschke]

I respectfully aknowledge your disagreement. I can only say in my defense that I was thinking of my future career with the Navy, and unfortunately for me, that led right through a polygraph. However, had I known the hell I was going to be put through for months and months on end, I probably would have done a tad bit more homework.

Likewise. Knowing what I now do about polygraphy, I would much rather have gone to my own pre-employment polygraph examinations informed and not ignorant, as I did.

Interestingly enough, in the NCIS prebriefing for the polygraph, they asked a group of us if we had done any research on the polygraph and specifically mentioned your website in a not-too-flattering light. Without mentioning you by name, they cast you as a disgraced and disgruntled person who has a bone to pick with the authorities.

Thanks for sharing this interesting tidbit. I've heard elsewhere that FBI polygraph examiners have falsely accused me of having "confessed to lying" during my FBI polygraph (I didn't, and I didn't lie), and a prominent polygrapher in Los Angeles has falsely portrayed me (and perhaps still does) as a fugitive from justice as part of his standardized introductory spiel.

It may not be unfair to say I'm "disgruntled" and have a "bone to pick with the authorities," but that doesn't mean I'm not speaking the truth about polygraph matters. For background about how I came to be a polygraph critic, see my public statement, "Too Hot of a Potato: A Citizen-Soldier's Encounter With the Polygraph."

Did the NCIS offer anything beyond an ad hominem argument against AntiPolygraph.org?