Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: xtnbreg and alternative methods


From   Mário Marques <mmarques@eeg.uminho.pt>
To   statalist@hsphsun2.harvard.edu
Subject   st: xtnbreg and alternative methods
Date   Mon, 10 Sep 2012 23:51:57 +0100

Dear Statalist-Users,

I need help about xtnbreg, fe. I am using Stata 11. My dependent
variable is a count of new subsidiaries for different countries-pairs.
I have read (from Allison and Waterman, 2002) that xtnbreg, fe is not
a true fixed effects model. Dependent variable is overdispersed and
there are within individuals time-invariant covariates.

Anyway I ran xtnbreg, fe, but I would like to compare it to another model.

One of the solutions (I do not know if it is correct) is run
xtpoisson, fe and adjust standard errors upward for overdispersion
using person's chi squared or deviance. Do you think this procedure is
correct? If so, how could I get the Pearson's chi square/deviance
statistic?

Do you know about other feasible alternative methods?

Thanks for your consideration,
Mario Marques
*
*   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/


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