George, 

I stand corrected.  I was writing from memory and did not look back to see who commented on a portion of the APA response or to recognize that you had already linked it.  It was actually a comment from Drew that was taken out of context:

Quote:


My point was that yes the APA did distance themselves from the report in this way, yet I saw little that could be considered irresponsible.  I saw little that seemed to be an attempt to debunk or discredit the report, and there was a seeming approval of the call for further research and development.  My question should have been directed to Drew, so I could get more specifics on what he disliked so I could re-read more critically.  

Lastly, George, the NAS report said validity was neither proven or disproven, we each could spin it either way, but I think it's a little extreme to spin it thus:
Quote:


When you get me a clairvoyant to have as much success as I have in resolving criminal investigations then I'll agree to your correlation.  Until then you might consider contributing spin material to James Carville.  

Regards

Public Servant,

You write in part:

Quote:


The point I'm trying to make is that if those making a positive assertion (e.g., polygraphers who assert that CQT polygraphy is a valid test for deception) wish to be believed, then it is incumbent on them to prove the assertion, and not on others to disprove it.

Such success as you may have had in resolving criminal investigations using CQT polygraphy may demonstrate its utility, but it does not provide any evidence for its validity.

Public Servant writes in part:

Quote:


I have certainly been critical of law enforcement's reliance on polygraphy in criminal investigations such as the FBI's ongoing anthrax investigation, or its abortive investigation into the matter of the missing hard disks at Los Alamos National Laboratory, or the Department of Energy and FBI's polygraph interrogations of Wen Ho Lee (see Chapter 2 of The Lie Behind the Lie Detector for critical commentary on the latter case.)

But I am not seeking to "eliminate any lie detection method available to the government in any form." No method of lie detection is known to exist. Instead, we have pseudoscientific nonsense such as CQT polygraphy and voice stress analysis, techniques that rely for their utility on public belief in the unfounded notion that they can detect lies. When a genuine method of lie detection is invented, I'll have no quarrel with it. But for as long as I am able, I'll continue working to expose CQT polygraphy for the fraud that it is. I believe that society will benefit when mention of a polygraph "test" evokes universal laughter, and the pseudoscience of polygraphy takes its rightful place next to phrenology and graphology in the trash heap of abandoned quackeries.

It's distressing to note the latest news story in the L.A. Times regarding many agencies still refusing to abandon polygraph screening.  I suppose it will take a bunch of lawsuits (possibly on due process/equal protection grounds?) or the actions of lawmakers to force security personnel to "do their jobs".

(That last remark really isn't intended to be as scathing as it sounds, BTW.  I really do understand why polygraph screening is used.  Security personnel, like everyone else, want to get the "most bang for the buck", and the most efficiency possible -- especially in an era of politicians who have sold the public on the idea that government services should just happen for free.  Unfortunately, polygraph screening simply doesn't work.  As they say, "you get what you pay for", and unfortunately, polygraph screening is cheap snake oil.  It's no substitute for the drudgery, expense and time investment of more thorough background checks.)

Skeptic

An especially relevant quote:

"No matter how disastrously some policy has turned out, anyone who criticizes it can expect to hear: 'But what would you replace it with?' When you put out a fire, what do you replace it with?" ~ Thomas Sowell

Public Servant:

I feel some (including you) are missing an important conclusion of the NAS regarding Specific Incident testing.  That is, results are "well below perfect."  To me, that means "well below 100%."   

So...maybe one quarter of the time, people telling the truth are falsely accused of lying?  That's certainly "well above chance" (50% accuracy), but would you like to be forced to take that chance on something that can really impact your life?

Now, here's another crucial tidbit that an LE officer should know.  POLYGRAPH EXAMS ARE NOT ALWAYS VOLUNTARY!!!  If you are on probation FOR ANY REASON, your probation officer may require you to submit to a polygraph.   

Although the results of a polygraph are not admissable in court, they are admissable as evidence in a probation violation hearing, where the judge must simply determine it is LIKELY that you violated your probation, not that you violated it beyond a reasonable doubt.   

Now, agreed, when one is convicted of a crime, one is justifiably forced to give up some of the rights guaranteed by our constitution.  HOWEVER, is being subject to the whims of unreliable ("well below perfect") science in determining our fate really consistent with this?

My fiance is on probation for a nonviolent crime.  He was falsely accused of threatening someone with a gun.  (He does not own or have access to a handgun, the accuser is a homeless con man that tried to extort money from us, and the guy served time a few years ago in CA for, coincidence, waving a flare gun in someone's face)

Rather than truly investigate, the PO required a polygraph, and my guy reacted w/more physiological stress to the relevant questions than to a poorly constructed control question. (Did you point a gun at the guy vs. Have you ever lied to law enforcement about a major crime?) The PO is going forward with a PV based on the poly and accusation alone.  (He did no further investigation).  My fiance faces prison time over this, which will assuredly destroy our lives and livelihood.  (and our ability to pay taxes and educate his sons - for you pragmatists out there)

Sure, we're just one sad anecdotal example of the  evils of frequent false positives.  but....

Answer me this:  Is society really better off with someone off the streets who may have pointed a gun at someone (a test with "Well below perfect" reliability said so)?    

In all honesty, if I had known about probation requirements and the unreliability of polygraphs, I wish I had sold my car and given the man the $8500 he attempted to extort from us.
Instead, we trusted the myth that the poly is accurate, and that "the truth will out". 

LE officers - is that really your interpretation of the constitution? If you mess up and find yourself on probation, you had better pay off anyone who tries to extort from you because if you don't, they could make a false accusation.  Then you'll be forced to take a poly, which you have say a 20-40% chance of flunking even if you're honest.

That's not my America.  IS it yours??

Our one chance is to perhaps take a 2nd poly.  It's been offered by a prominent academic.  However, I know my guy well.  With all that we'd have riding on it, I can't fathom a control question that would stress him out more than "Did you point a gun at that guy?"  (A requirement for "passing" the test.)  How can I trust a guy who doesn't even know him to come up with one?

sie,

Answers to the questions you raised:

1 & 2) Regarding what the GKT (Guilty Knowledge Test, also called a "Concealed Information Test") is and how it works, you'll find it briefly explained in Chapter 3 of The Lie Behind the Lie Detector. A much more detailed explanation may be found in the 2nd edition of David T. Lykken's book, A Tremor in the Blood: Uses and Abuses of the Lie Detector (Plenum Trade, 1998). Here is the NAS report's explanation, cited from Appendix A, p. A-4:

Quote:


3) Yes, the GKT/CIT can be manipulated by either the examiner or the examinee.

4) The GKT is being employed by law enforcement, but not to a great extent.

It's doubtful that a meaningful Guilty Knowledge Test could be constructed for the behavior of which chicbette's fiance stands accused.