RE: st: Beta coefficients for GLM models?

["Scott Holupka" <Scott.Holupka@jhu.edu>]

Thanks.  You are correct, I'm running a GLM with a gamma-distribution and
log link.  

So does that mean that I should treat the coefficients directly as percent
change values, or do I need to take the exponent, as I believe is normally
done with logit models?

Thanks again for the help.

Scott




-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Austin Nichols
Sent: Tuesday, November 20, 2012 9:18 PM
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: Beta coefficients for GLM models?

Scott Holupka <Scott.Holupka@jhu.edu>
The crucial information here is which link you are using, but assuming
a log link, the coefficient is essentially a percentage change in
expenditure; if your RHS variables are logged, you get an elasticity,
which is scale-free.

On Tue, Nov 20, 2012 at 4:59 PM, Scott Holupka <Scott.Holupka@jhu.edu>
wrote:
> I'm currently using GLM (Stata 12) to analyze some expenditure data and I
> would like to compare the effects of different coefficients in the model.
> If this were an OLS I would look at the beta coefficients, but I can't
> figure how to compute a beta or beta-like coefficient for a GLM model.  Is
> there any way to do this?
>
> Thanks for any suggestions or advice.
>
> Scott Holupka
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/