st: correlation table for fixed effects negative binomial model

[Mukund Chari <mukundc@uw.edu>]

Dear Stata users,


 I am trying to get the correlations matrix after running a fixed effects negative binomial model (xtnbreg). I cannot use the mkcorr command because it uses all the observations in the dataset, while the xtnbreg command drops observations for the groups for which it finds no variance, thus resulting in a mismatch on the number of observations between that of the regression and the correlation routines. Since I have a very large number of fixed effects (30000), I cannot use an unconditional fixed effects model with dummies as it would be computationally challenging.

 To get to the "true" correlations table, I could use the estat vce, corr   command after running the regression, but this would only return the estimates, and not the significance levels. I tried searching through the forum for appropriate solutions, but could not find any discussion on this topic.

 Any help would be much appreciated.

*This is my first post to the forum, please excuse any inadvertent violation of forum norms*

Best,
Mukund Chari.




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