Re: st: AW: dropped groups in xtlogit fixed effects

[rdae20@bath.ac.uk]

I have data that covers 30 years and 90 countries.

For about half of those countries the LHS variable is 1 just once, for  
about a quarter it is once more than once and for the rest it is zero  
for the whole period. The RHS variables all vary significantly within  
and across countries.

Within country variation is low for the LHS, but not the RHS, what  
would you recommend? I am willing to sacrifice the 25% of data (is  
always equal to zero) for better results.

Many thanks,
Richard

Quoting Maarten buis <maartenbuis@yahoo.co.uk>:

> --- On Thu, 1/4/10, rdae20@bath.ac.uk wrote:
> > So xtlogit, fe only measures within country variations right?
> >
> > If I want to measure both cross country and within country
> > what should i use? random effects?
>
> If you want to look at variables that are (virtually) constant
> within countries but differ across countries, then you can't
> use a fixed effects regression, and you'll have to use a
> random effects regression.
>
> -- Maarten
>
> --------------------------
> Maarten L. Buis
> Institut fuer Soziologie
> Universitaet Tuebingen
> Wilhelmstrasse 36
> 72074 Tuebingen
> Germany
>
> http://www.maartenbuis.nl
> --------------------------
>
>
>
>
> *
> *   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/




*
*   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/