[tips] RE: Nonparametric Effect Size and Post-Hoc Power

[Christopher D. Green]

I cannot tell you how to calculate the power of nonparametric tests, but I am certain that one can, because there is an extensive scholarly literature comparing the relative powers of nonparametric tests to their parametric "cousins" (in which it was initially argued that nonparametrics were much less powerful -- which is why we haven't traditionally used them as much as parametrics -- and later finding that even fairly minor deviations from assumptions lower the power of parametric tests enough to make nonparametrics highly competitive -- which is why there is now renewed interest in them). Regards, -- Christopher D. Green Department of Psychology York University Toronto, ON M3J 1P3 Canada 416-736-5115 ex. 66164 [EMAIL PROTECTED] http://www.yorku.ca/christo ============================= Marc Carter wrote: Seconded.   I'd be interested in hearing about this, too.  It seems to me that the computation of power has to make assumptions about the shape of the distribution of the dependent variable (power is essentially a measure of area of the distribution of the variable -- under the alternative hypothesis -- above the criterion), and so if we cannot make assumptions about the character of that distribution (that's why they're called "distribution-free stats"), I'm at a loss to figure how we'd compute its area.    I'm wondering if there's some way to bootstrap a distribution based on the data, generate a function to describe it, and then get about integrating it.   But, as often happens, I could be wrong and would really like to know.   m ------- "Mauchly's Test of Sphericity: Tests the null hypothesis that the error covariance matrix of the orthonormalized transformed dependent variables is proportional to an identity matrix." --- SPSS   From: ROBERT [EMAIL PROTECTED]@MATHSCIENCE [mailto:[EMAIL PROTECTED]] Sent: Thursday, September 28, 2006 6:47 AM To: Teaching in the Psychological Sciences (TIPS) Subject: [tips] Nonparametric Effect Size and Post-Hoc Power Anyone out there familiar with way to calculate effect size and post-hoc power following nonparametric analyses (specifically Kruskal-Wallis ANOVAs and Mann-Whitney U tests)?   Thanks,   Rob Flint -------------------------------------------------------------------- Robert W. Flint, Jr., Ph.D. Associate Professor of Psychology The College of Saint Rose 432 Western Avenue Albany, NY  12203-1490   [EMAIL PROTECTED] 518.458.5379   Behavioral/Cognitive Neuroscience Web Page - http://academic2.strose.edu/Math_And_Science/flintr/   Journal of Behavioral and Neuroscience Research - http://academic2.strose.edu/Math_And_Science/flintr/jbnr/   --- To make changes to your subscription go to: http://acsun.frostburg.edu/cgi-bin/lyris.pl?enter=tips&text_mode=0&lang=english --- To make changes to your subscription go to: http://acsun.frostburg.edu/cgi-bin/lyris.pl?enter=tips&text_mode=0&lang=english --- To make changes to your subscription go to: http://acsun.frostburg.edu/cgi-bin/lyris.pl?enter=tips&text_mode=0&lang=english