Hi Martin,
thanks a lot, works great! The first colon was a typo. The actual
model I am trying to estimate is quite a bit more complex than the
"empty" one presented. so far xtmixed works fine for me.
Regards
Christian
On Thu, Sep 17, 2009 at 5:27 PM, Martin Weiss <[email protected]> wrote:
>
> <>
>
> BTW, are you sure about the first colon after the double pipe symbol?
>
> If all you want is the model you have presented here, you could get it more
> easily from -xtreg- (where results are more intuitively accessible):
>
>
> ********
> clear*
> vers 10.1
> set obs 1000
>
> //id
> gen id=_n
>
> //individual effect
> gen alpha=rnormal()
>
> //expand to # of time periods
> expand 10
>
> //get time period (after expansion!)
> bys id: gen time=_n
>
> //residual per time period
> gen eps=rnormal()
>
> compress
>
> //xtset the thing
> xtset id time
>
> //dependent
> gen mydependent= /*
> */ rnormal(3,10)+alpha+eps
>
> xtmixed mydepend || id:
>
> xtreg mydepend, re
> ********
>
>
>
> HTH
> Martin
>
>
> -----Ursprüngliche Nachricht-----
> Von: [email protected]
> [mailto:[email protected]] Im Auftrag von Christian Weiß
> Gesendet: Donnerstag, 17. September 2009 23:03
> An: statalist
> Betreff: st: Storing random intercept estimation? (xtmixed?)
>
> Dear Statalist,
>
>
> using a textbook example on birthweight I estimated the following
> model: xtmixed birwt ||: momid:, variance
> yielding
>
>
> ----------------------------------------------------------------------------
> --
> birwt | Coef. Std. Err. z P>|z| [95% Conf.
> Interval]
> -------------+--------------------------------------------------------------
> --
> _cons | 3467.969 7.138554 485.81 0.000 3453.977
> 3481.96
> ----------------------------------------------------------------------------
> --
>
> ----------------------------------------------------------------------------
> --
> Random-effects Parameters | Estimate Std. Err. [95% Conf.
> Interval]
> -----------------------------+----------------------------------------------
> --
> momid: Identity |
> var(_cons) | 135686 4755.874 126677.6
> 145335
> -----------------------------+----------------------------------------------
> --
> var(Residual) | 142625.4 2965.965 136929.1
> 148558.7
> ----------------------------------------------------------------------------
> --
> LR test vs. linear regression: chibar2(01) = 1316.12 Prob >= chibar2 =
> 0.0000
>
> with "di _b[_cons]" I can display (and thus process) the estimated
> coefficient of birwt. However, I would like to process the estimated
> variance components of the random intercept (i.e. var(_cons) and
> var(Residual).
>
> How can adress these values?
>
> Best,
> Chris
>
>
>
>
> Viele Grüße
> Christian
>
> *
> * 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/
