In a message dated 10/31/02 5:05:29 PM, [log in to unmask] writes:

>You could make
>a case that every group that does studies on anything has a vested interest
>in proving one outcome or another... so why are you confident in statistics
>that you agree with, but not those that state the opposite of what you
>believe?

As the NRA says, "Guns don't kill. People kill." The same is true for
statistics. Stats don't distort the facts, statisticians do. And, yes, there
are bases even in "pure sciences", like physics, the experimenter has a
measurable effect on the results in some cases. In an ideal world, an
investigator would objectively select methods and statistics which were
completely objective and use sampling which was without bias, but this is not
and ideal world!

Bobby