Re: st: xtmixed to estimate the effect of id and time t

[Song Nhac <songnhac@gmail.com>]

Thank Marteen,

The command should be: xtmixed y x1 x2 || _all.R:id || _all.R:t

But because of unbalanced data, the command cannot caculate the
gradient. I try to use other methods: with technique (bfgs) or put
more option: mle
but I cannot get the standard error when perfoming the command.

Thanks,
SN

On Mon, Sep 12, 2011 at 2:13 PM, Maarten Buis <maartenlbuis@gmail.com> wrote:
> On Mon, Sep 12, 2011 at 2:07 PM, Song Nhac wrote:
>> I have an unbalanced data set with id is firms' id and time t, I would
>> like to measure the effect of unobservable part- but the random effect
>> command just assumes that time-invariance and predict results based on
>> the firms' id. So if I want to measure the effect of both id and time,
>> I could use the command xtmixed? it will allow the effect vary by time
>> or not?
>>
>> My command is:
>> xtmixed y x1 x2 || id: || t:
>
> That is not right as t is not nested in id. What you seem to want is a
> so called crossed effects model. In the manual there is an entry on
> how to estimate such a model.
>
> Hope this helps,
> 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/