Slashdot Mirror
Veteran FBI Employee Accused of Trying To Beat Polygraph, Suspended Without Pay
George Maschke writes: A mid-career veteran of the FBI has been suspended without pay and faces revocation of his/her security clearance (which would inevitably lead to termination) because the Bureau's polygraph operators allege he/she tried to beat the polygraph. The case is currently the subject of an unpublicized Congressional inquiry. Retired FBI scientist, supervisory special agent, and polygraph critic Dr. Drew Richardson has publicly shared a memorandum he wrote in support of the accused in this case, which has heretofore been shrouded in secrecy. It should be borne in mind that polygraphy is vulnerable to simple countermeasures (PDF, see Ch. 4) that polygraph operators cannot detect. This case is yet another example of how the pseudoscience of polygraphy endangers virtually everyone with a high-level security clearance.
When there is overwhelming evidence that polygraphs don't work?
You probably already know this, but for those who may not, the polygraph is mostly an interrogative tool used in eliciting confessions or telltale behavior regardless of its real ability to gauge honesty. As for "beating" a polygraph, the charge is as spurious as the basic claim that it can gauge honesty. If it can't, and it's largely been demonstrated that it can't, there's no reason to hold anyone to the results it presents regardless of what the operator may believe they indicate.
Penn & Teller "Bullshit!"
http://www.220.ro/emisiuni-tv/...
"60 Minutes"
https://www.youtube.com/watch?...
I had a personal friend who has a PhD in Engineering who worked for an US govt agency with a 3-letter abbreviation. He got so fed up with the idiocy of periodically put put on a polygraph, that he quit. It seems every time it happened, they would come up with yet another bogus accusation, and try their damnedest to get a confession.
I the city that I left (due to government corruption, a failing economy, and did I mention a collapsing economy) hired the brother of the premier to run the polygraphs for the fire and police departments. But it wasn't corruption at that level of simplicity, police chiefs and what not all had their fingers in the pie. I don't think if any of them care if the things work all they care about is getting their contracts and feathering their beds.
I suspect that if you go to the FBI they have whole divisions that have been build upon the foundation of polygraph technology. They not only would suddenly have to actually be FBI agents but they all would have just spent the bulk of their careers basically interpreting goose livers. But even worse they would not be able to go out on consulting gigs where they can be the "FBI Polygraph Expert" this would be a total disaster.
Lastly I suspect there is a bit of powertripping among their numbers. You can point to some squiggles on a line and say, "His answers were weak, here, here and here." and you have just ruined a career or sent a person to jail.
So it doesn't really matter how many times the FBI is shown to be using science at the level of a cave man witchdoctor they have a massive PR machine plus their argument trumps your argument because they carry guns and can lock you in a cage for disagreeing.
I have a family member who applied to the Border Patrol. They are now giving polygraphs to all applicants. After failing 4 polygraphs for stuff like "did you kill somebody and bury them in a shallow grave" to "are you part of a criminal organization" and "are you providing materiel support to terrorists" he was dismissed from consideration. I assure you, they none of those were true, but this person is a very bad liar and gets nervous when accused. Border patrol currently have something like an 80% failure rate on pre-employment screenings. What should be particularly frightening is that we are actually selecting for liars who don't get nervous. The polygraph proponents will vehemently argue it can't be beat, which is technically true. It detects what it detects, pulse, respiration, blood pressure, skin galvanometer. But of course it's the meaning that makes the difference isn't it? It's not a lie detector, it's a nervousness detector. Do you really want people who don't get nervous when questioned?
Also, the polygraph is an excellent example of "base rate fallacy", and almost certainly the vast majority of people caught by the polygraph are completely innocent of anything. Even if the polygraph is 80% accurate (wildly generous) that means that if you test 10,000 national security employees you are going to fail 2000. How many spies do you have? A reasonable reasonable estimate would be in the single digit range, but let's just say it's 20. That means you are going to "catch" 2000 good employees and only 8 spies. 0.4%, which is going to make prosecution virtually impossible. You are not even going to be able to devote much investigative effort since 99.6% of the time it is going to end up being a waste of time.