Hello Statlist,
I am doing meta-analysis using various routines like < metareg > for
weighted random effects regression models.
I need to use the b coefficients to estimate semipartial correlations
between Y and X1 with one covariate (X2) partialled out. A literal
example is:
Y = antisocial behavior
X1 = social self control
X2 = initial group equivalence (potential methodologiocal
confound)
I need the semipartial correlation for Y and X1 to see if their
relationship changes after controlling for methodological confounds like
X2.
I know how to directly compute a semipartial corr from ordinary
bivariate correlations. But, in this case I need the semipartial corr to
reflect the weighted random effects (e.g., max likelihood) models used
in meta-analysis. I guess one way to do this is to do separate
bivariate weighted random effects regressions for every pair of
variables I am interested in, and then use the bivariate b to manually
compute the semipartial correlations. But I need to work with a lot of
variables and am hoping there is a better way.
So, does anyone know a straightforward way to use regression
coefficients to calculate semipartial correlations in Stata? I think I
have seen this done before, but can't find references or examples.
Thanks very much,
wg
*
* 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/
