wrote on Sep 12^th, 2003 at 5:46pm:

No one is arguing here that props (including the polygraph) cannot be useful in interrogations. But as the National Academy of Sciences noted in its report, there is essentially no evidence on the additive validity of polygraphy (its ability to add predictive value to that which can be achieved by other means).

While true, predictive value is not established for many qualification standards. OTOH, one can argue establishing predictive value should be a precondition for highly invasive processes as the polygraph.


Quote:

And as Dr. Fienberg pointed out, the handful of purported successes of the polygraph that Kyle McSlarrow adduced are of no statistical value.

OTOH, I think it is obvious that if such information was available, it would not be prudent to publish it's detail as it would be valuable for adversaries.


Quote:

Polygraph screening is a fraud, as more and more of us who are or have been subject to it are learning. As Dr. Richardson has pointed out before, continued polygraph screening would require such a universal bluff as to be impractical over the long term.

Perhaps this is true over the long term, but it would appear that the new information discovered  from admissions remains at a fairly high level and so the "long term" may be much further out. Admission rates are likley a good proxy for the effectiveness of the polygraph "bluff."


Quote:

Do you agree with Kyle McSlarrow's stated plan for continued reliance on polygraph screening?

I don't know enough about it to make that determination assuming a purely utilitatian approach. OTOH, I have a very visceral reaction to the idea of false positives from a screening test that is deceptive on it's face. It appears Kyle has adopted a utilitarian approach and has factored in consequences such as loss of talent, etc.

-Marty