I am sorry for not having been precise enough. What I would normally
need is -xtlogit-, not xtreg. And, if I am not wrong, -xtlogit- won't
allow me to cluster individuals on a higher level (households, in this
case). Best, Dimitrije.
2010/1/2 Martin Weiss <[email protected]>:
>
> <>
>
> If all you want is a random _intercept_, you should of course look at
> -xtreg, re-, which does allow a -cluster()- option.
>
>
> HTH
> Martin
>
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of Dimitrije Tišma
> Sent: Samstag, 2. Januar 2010 16:29
> To: [email protected]
> Subject: st: cluster option and random effects
>
> Hi Statalisters,
>
> I have a very short yet possibly tough question. I need to estimate a
> random intercept model using data with not entirely independent
> observations within households. Basically, I would like to know what
> possibilities there are in Stata to estimate random effects when using
> cluster( ) option. I have already used -gllamm-, but the coefficients
> of duration dependence variables do not the way I would expect them
> to. So, I wanted to ask if there are there any alternatives to
> -gllamm- (which I like and use a lot, but I really need to be
> comfortable about my results and want to do double check)? Happy new
> year to all, Dimitrije.
> *
> * 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/
>
*
* 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/
