Gordon Barland wrote:


Quote:

There may be a problem with semantics here. I did not mean confirm in the sense of absolute ground truth. That is extremely difficult to do in any real world situation. Therein lies the big advantage of laboratory research, where we made our initial breakthroughs in detecting mid- and high level countermeasures. What I did mean was confirm to the satisfaction of the individual agency. The steps that must be accomplished to reach that level of confirmation vary from one agency to another. The key point is that in the Federal government, most examiners do not make such a decision arbitrarily. They must follow their agency's guidelines, which may require concurrence from their quality control office.

Gordon, confirmation "to the satisfaction of the individual agency" is no real confirmation at all, is it? In your first post, which started this message thread, you wrote:


Quote:

Each Federal agency has a quality control office which reviews every examination. I know of a number of cases where the examiner cleared the person taking the test, but the quality control office detected apparent countermeasures. The persons were re-examined, and the use of countermeasures was confirmed.

All you really meant to say then, is that some federal agencies will decide (based on undisclosed criteria that vary from agency to agency) that a subject employed countermeasures, whether or not the subject admits to it?

What evidence exists that DoDPI-trained polygraphers (or, indeed, any polygraphers) can detect the countermeasures described in The Lie Behind the Lie Detector at better than chance levels?

Lastly, if you would use the relevant/irrelevant format with sophisticated subjects (i.e., those who understand the polygraph procedure), then on what scientific basis do you expect to be able to distinguish truth from deception using this (thoroughly discredited) technique? For the informed, truthful subject who heeds your advice and does not employ countermeasures but instead admits to his/her knowledge of the trickery on which "control" question "test" polygraphy depends, the promise of being treated to a relevant/irrelevant "test" instead is hardly reassuring.