On 30 Sep 2002 12:11:34 -0700, [EMAIL PROTECTED] (Michael Granaas) wrote: ... > I have a student who wants to test for differences in effect sizes
" ... differences in effect sizes" could refer to an interaction, if "effect size" refers to main effects -- > across > several measures for two different groups. " ... across several measures" is not too difficult, I expect, for *parallel* measures. > I've never heard of such a > test and wondered if such a thing exists. Any references? Some interaction in some Repeated Measures? I like to draw up explicit contrasts, after I know what the hypothesis is. As of yet, I don't think I grasp your student's hypothesis, which seems undefined rather than impossible. -- Rich Ulrich, [EMAIL PROTECTED] http://www.pitt.edu/~wpilib/index.html . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
