Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: RE: can I compare correlations with t-tests (Bonferroni)?


From   "Verkuilen, Jay" <JVerkuilen@gc.cuny.edu>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: RE: can I compare correlations with t-tests (Bonferroni)?
Date   Mon, 16 Jun 2008 12:05:32 -0400

>>Tests of correlations are notoriously sensitive to the normality
assumption.  So any solutions may be subject to that problem.  One
approach might be to use the Fisher z transformation for the
correlations and then test to see if each is 0 or not.  I'd be cautious
about any p-value citations. To compare correlation A1 with A10 you
would use z(A1)-z(A10) and go from there.<<

Yes, be wary. 

Fisher Z helps as differences are much more linear on the Fisher Z
scale. 

You may also want to set the problem up as an SEM and use SEM software,
e.g., the free MX by Mike Neale. http://www.vcu.edu/mx/ If your sample
size is large, you could use AWLS.

Bootstrapping is another possibility. 


*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index