Quote from: SanchoPanza on Jan 03, 2008, 09:16 PMQuoteKeep in mind that this was a screening exam study. By their very nature, screening exams should be designed to make more of those errors in order to reduce the possibility of a false negative. We can recover from one, but not the other. Moreover, reactions in screening exam are (or should be) only considered tentatively positive (meaning you're not done).
So, you are saying that it acceptable that a "screening exam" only be 75% accurate? That it is okay to brand 25% of police applicants liars, knowing that the false positive polygraph result will follow them throughout your carreer?
And you declare that you are a Man of God?
"A lie consists in speaking a falsehood with the intention of deceiving. The Lord denounces lying as the work of the devil: "You are of your father the devil, . . . there is no truth in him. When he lies, he speaks according to his own nature, for he is a liar and the father of lies."
QuoteAccuracy rates from laboratory studies with volunteers committing mock crimes and being polygraphed in the absence of jeopardy, such as the 2007 study by Honts and Amato to which you refer, cannot be expected to generalize to field conditions, where examinees face potentially severe consequences for being (rightly or wrongly) deemed deceptive.
QuoteBut in any event, in the Honts & Amato study, only 63% of innocent examinees were correctly judged truthful by human examiners, while with automation, 76% of innocent examinees were judged innocent. In both cases, a substantial percentage of innocent examinees failed to pass. Under field conditions, one would expect even worse results.
Quote from: karasoth on Dec 30, 2007, 09:31 PMI disagree. In the Griebl Case the FBI was given ample information that Griebel had showed deception on his examination and then failed to give it proper weight. That is evident in the record you provided.
QuoteQ.--Are you double-crossing the agents?
A.--No.
Q.--Are you sincere in present efforts to assist Federal agents?
A.--Yes.
QuoteIn the other traitor case, Investigators were again given information regarding his deceptive responses to relevant issues and again chose not to give the information proper weight. The investigators were so worried that the Traitor would find out he was under investigation that they also deliberately withheld information that would have been relevant to any examiner who was trying to determine if he was a spy.
QuoteHowever you choose to characterize the document YOU provided or these official records the language is plain. Neither of these traitors beat "Polygraph" they just beat the investigators.

QuoteThis procedure has no scientific basis whatsoever, and large numbers of false positives are to be expected.
Quote from: karasoth on Dec 30, 2007, 05:03 AMSancho Panza,
According to Weiner, Johnston, and Lewis, the questions Ames was asked were the standard questions he and everyone else were always asked, specifically:QuoteHave you divulged any classified information to any unauthorized person? Have you had any unauthorized contact with foreigners? Have you gone to work for the other side? Have you been pitched--that is, approached--by a foreign intelligence service.
Regardless of what information the polygraph examiner did or did not have, there is no doubt that Ames was lying on the first three questions. But neither the polygraph nor its operator was able to discern that. It seems you are suggesting that, had the operator known in advance that the subject would be lying, he or she would have had a much better chance of accurately detecting deception. While that is almost certainly accurate, it is hardly a ringing endorsement for the accuracy of the polygraph.
The final question was actually answered honestly, since Ames had approached the Soviets, not the other way around. But that is only answer the examiner said indicated deception, and after Ames gave his set explanation the conclusion was as follows:QuoteThe polygraph operator deemed Ames forthcoming in all respects, and he called Ames's responses "bright" and "direct."
I don't see how this incident can be seen as anything other than a typical example of the inability of the polygraph to detect deception.