Skeptic wrote on Sep 12^th, 2003 at 11:46pm:

Of course.  But it also doesn't mean such basis does exist, either, and since the people in question want to use the polygraph and (in some cases) have a vested interest in it, the burden is upon them to prove that utility, rather than upon others to prove the negative.

I agree. In fact it would be quite impossible to prove a negative in this instance.  That said, there is no obligation (though it would certainly be desireable) to disclose any proven utility to the outside.


Quote:

OTOH, letting people know that you have a padlock on your door, and that its a 3" Masterlock made of hardened steel, and that 10 burglars have tried without success to open the lock, would likely have a deterrent effect.  ... ...  If you are dealing with an adversary who is sophisticated enough to scientifically structure an effective breach of security, you're likely dealing with someone who is capable of doing their own studies on polygraphs and knows the weaknesses full well.  If this is the case, failing to release not just full details but *any* details regarding how the polygraph has or has not caught spies only serves to keep ourselves in the dark regarding its utility.

In general, any information one discloses to an adversary, about whether something works or not, is useful to allow refocussing resources. As for the more capable adversarys, knowing the mechanics and weaknesses of polygraphs is a long way from knowing the kind of statistical information that is lacking here.

That said, I do believe and I think anecdotal and published evidence supports, some types of countermeasures, especially combined with practice on a polygraph, could be very effective. Enough so as to effectively limit the main use of the poly to a prophylactic utility. It likely is far less effective against well trained moles. Since we have caught some of these spys, it would be interesting to learn what training they recieved re polys. Don't think we will ever see that in public though.

-Marty