05-14-2012 02:07 PM
Can someone explain me how I can determine the overall model fit for a negative binomial and Poisson model?
(I'm not an expert with SAS enterprise guide, so please explain it simple :-) )
I want to know how much of the variance is determined by my estimated models (in order to check if it's a steadfast model
Thanks in advance,
05-15-2012 09:11 AM
While I cannot offer specific analytical advice (not my background), your request seems to fit the use of PROC GENMOD in SAS. To find this in SAS Enterprise Guide: Tasks->Regression->Generalized Linear Models.
That's a place to start. I don't know if it will take you all of the way to your solution.
05-15-2012 12:42 PM
what I actually mean is;
how much of the variance is determined by the model.. I thought that normally the r² calculated this rate but my NEGBINOM and POISS did not calculate this..
I want to know how much of the variance is determined by my estimated models (in order to check if it's a steadfast model)
can someone help?