sackett wrote on Mar 11^th, 2008 at 7:59pm:


While other have made the coin-toss analogy, I don't think it is a particularly helpful one for understanding the variables that are at play in the conduct of a polygraph examination. The likelihood of a coin toss coming out heads or tails can be determined with precision. The likelihood of any particular polygraph "test" result being correct or incorrect cannot be determined with such precision.

Quote:


The difficulties in establishing ground truth under field conditions is an entirely separate matter from the lack of standardization and control that besets polygraphic lie detection. There is only one polygraph technique that can truly help to determine if a person is withholding information (absent a corroborable confession), and that is concealed information testing (such as Lykken's Guilty Knowledge Test). The much more widely used Control Question Test cannot determine whether a person is withholding information.

Quote:


No mantras, Sackett. But some truths bear repeating. Especially in the face of repeated mis-statements from polygraph operators.

George W. Maschke wrote on Mar 11^th, 2008 at 7:44pm:


Oh come on George.  You wouldn't begrudge me a little literary license and interpretive summation, would you?  Not unlike many statements made here by some of your followers, would you?

And you're right.  Many times ground truth is not knowable; however, when someone is withholding information, lying about something, etc, etc, the polygraph procedure works very well to identify it.


Sackett

P.S.  Remember to insert the standard mantra in your reply.  I would truly be disappointed if you forgot.  It's actually kind of fun to read...

sackett wrote on Mar 11^th, 2008 at 7:30pm:


I haven't stated that polygraph testing is like a coin-flip (that is, a 50-50 proposition). Rather a meaningful accuracy rate for polygraphic lie tests is not truly knowable, because the procedure lacks standardization and control. Note also that digithead did not state that polygraph testing has an accuracy rate of 54%. That figure is the base rate of deception in the study population.

digithead wrote on Mar 11^th, 2008 at 4:41pm:


I am not a researcher and certainly not an academic; however, I do know many variables play into the validity and reliability of any research project.

While I didn't read the entire research report nor study the valididty of it, I wonder, who were the 74 people?  Were they college student just back from spring break with stories to tell? were they working people with problems of their own and given three minutes of quiet would focus on them.  Could this have caused labile responses during this "evaluation" time.  Possibly misdiagnosing otherwise stabiles as labiles.  I don't see any references to control group by which to determine the accuracy of the results.  I see no randomness, which I'm sure you know is the key to many research projects.  I certainly didn't see anything remotely resembling reward or fear of consequences (as in real life), and as is imperative to polygraph research; otherwise, who cares if you catch me or don't... You didn't report the hypothesis, which as you know if not guarded against, becomes the outcome through self proclaimed prophesy, versus actual findings, etc, etc, etc.

I'm not suggesting this project was done incorrectly, but before I make any opinion about it, I want someone with authority, expertise and academic creditials to review all aspects and method to the results to verify it's validity.  There are many, many "research" projects out there which are worth nothing more than a quick publishing.  I would never blindly assume the one you presented here, or which anyone might present, is automatically accurate and valid until reviewed by someone who has that ability.   

On the lighter side, you disproved George's assertions that polygraph is as good as a coin flip.  54%!  It's better!  Thanks for helping...

Sackett

Quote:


The FFF response can be aroused without a full blown response.  IOW, your BP goes up slightly...etc.

But here is something I don't understand.  The FFF response is controlled by the lower brain and unconscious mind.  Typically, the polygrapher, after stopping the tape, and informing the test subject that they "had trouble with the question on....", will ask "WHY ARE YOU HAVING TROUBLE WITH THIS QUESTION?"

But the conscious mind is not really involved!  How the hell would one know why their autonomic nervous system got just got aroused.  This would explain why people get totally perplexed when told they are being deceptive.  They are being asked to consciously figure out why an unconscious process in their body has reacted (in the polygraphers parlance "Why are you lying!").

Another key point regarding FFF.  It puts one on guard to react to PERCEIVED THREATS, not necessarily REAL THREATS.  It is a primitive response and according to Dr. Herbert Benson, in his book "The Relaxation Response" the FFF response is often elicited in modern man without any REAL threat, and proper avenue for release.  This causes much undue anxiety and stress.

And the pre-employment "witch hunt" test is very broad and arbitrary.

Whoops, gotta go.  My BP is up, and my hands are slightly moist.  I have to go into my meditation room (latrine) and search my soul for what can be causing that!   

Quote:


In your dreams.

Quote:


He said it on the "Oreilly Factor" (FNN) about two months ago with millions of people watching.

Polygraphers routinely make such bold statements to the gullible public.  Funny they don't on this board.

sackett wrote on Mar 11^th, 2008 at 3:34pm:


I'm off to the airport so I won't be responding to whatever your next thing is but a quick search of Pubmed turned up this gem from your own polygraph people in 1981, enjoy!

Waid WM, Wilson SK, Orne MT. 1981. Cross-modal physiological effects of electrodermal lability in the detection of deception. J Pers Soc Psychol. 40(6):1118-25.
Quote:


Bolding mine. Btw, labile means unstable...

At least I finally found research that shows that increased individual variability in the four channels measured by the polygraph affects the polygraph's ability to detect deception...

I'll add here that the probability of deception (i.e. base rate) in this sample is 54% which is hardly ever the case in a screening situation (it's usually much lower). This just adds to the evidence that the polygraph is biased against the truthful, especially amongst with those having unstable physiological responses...

Regards...

T.M. Cullen wrote on Mar 11^th, 2008 at 3:47am:


I don't know, I'm not a researcher, I'm just a lowly practitioner.  What does the research results indicate?  80-95%??  If so, then how is it that Trimarco (who BTW I have never actually heard him say that) would be wrong, IF, he was simply taking the higher end results?   

But of course, George often reminds us that it is all a flip of the coin, based on "the biblical" text of the NAS study...

Sackett

digithead wrote on Mar 11^th, 2008 at 3:33am:


Got any peer-reviewed research to show that these things do affect the CQT polygraph?

The first few are cognitive thought issues, contributing to the overall state of general nervous tension through the central nervous system,  the rest are biologically altering issues.  The severity one can only assume.  Without having a "specimen" in the chair, one can only hypothesize how or even if their disorder would effect the accuracy of polygraph.  I suspect, unless they are dysfunctional in nature they would not.   

I have tested many people with anxiety disorders, diabetes, hypertension, etc and have gotten good results.    

Sackett

The Discovery Channel program to which T.M. Cullen refers is episode 3 of Human Body: Pushing the Limits. Information about this program, including video clips from various episodes (including the polygraph segment), is available here:

http://dsc.discovery.com/tv/human-body/human-body.html

I've always been perplexed when polygraphers speak of the "fight or flight" response (or, alternatively, the "fight, flight, or freeze" response, sometimes abbreviated F³ in polygraph literature) in connection with polygraphy. When I sat for my polygraph examinations, I never had the urge to strike the polygrapher (fight). Similarly, it never occurred to me to flee the scene (flight): were I so disposed, I wouldn't have sat down in the polygraph chair to begin with! Similarly, the thought of remaining silent and motionless (freeze) didn't cross my mind either.

So if any of our friends in the polygraph community could explain how this "fight or flight" response has anything to do with the situation of an examinee in the polygraph suite, I'd be interested in learning more about this notion.

Quote:


Is the polygraph 95% accurate as he claims?

sackett wrote on Mar 11^th, 2008 at 3:16am:


Hmm, disgust, anger, nervousness, undetected hypertension, vascular diseases, diabetes, thyroid problems, or endocrine disorders immediately come to mind...

Got any peer-reviewed research to show that these things don't affect the CQT polygraph?

T.M. Cullen wrote on Mar 10^th, 2008 at 6:25pm:


Mr/Ms Cullen,

I saw the same program (accidentally flipping, don't you know).  I thought they explained the polygraph process fairly well.   

Once again, it doesn't fit the mantra of this site, but truly, can you name any reasons why a person would elicit an ANS response to a well reviewed and rehearsed question in a controlled setting, other than the dichotomy of developing a lie (or withholding, minimizing, secreting or rationalizing previously provided information)?

Sackett

P.S.  What exactly are you looking for with your question: "Why won't the polygraphers on this board support Trimarco in his claim?"   

The Discovery Channel aired a show last night about the human body.

They were talking about the human nervous system.

They featured the polygraph in a segment about the autonomic nervous system, and the "fight, flight or freeze" response.  Which is what is measured on the polygraph.  The show gave the impression that people who lie will often elicit this response.   Which is true.   

But they didn't bother to mention that people who ARE NOT LYING can elicit the same response when asked a question if they consider threatening.  Is the act of lying the only reason a person could consider a question to be even slightly alarming?

If you ask a father:  "Did you have sexual relations with your daughter last night?"  Do you think most fathers would have an ANS reaction?  Does that mean they are lying?

They show Trimarco testing a hispanic looking male who was sweating profusely.  Probably had him doing jumping jacks prior the filming.

Isn't Trimarco the guy who claimed on national TV that the test is 95% accurate?

Why won't the polygraphers on this board support Trimarco in his claim?   

TC