Harold W Kerster <[EMAIL PROTECTED]> wrote: > Yes, a t test may be enough. Students should also learn that results > obvious on their face need no stat analy at all -- don't gild the lily.
And then they run into a radical social-contructionist committee member who insists that the observed difference in pregnancy rates between the male and female subjects is not "obvious on its face" and demands a p-value :) . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
