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]

Re: st: XTReg re/mle clustered & weighted models


From   "Castle, Melissa H" <mhc14@pitt.edu>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   Re: st: XTReg re/mle clustered & weighted models
Date   Tue, 21 Dec 2010 14:13:04 -0500

Hi Stas (& stata-listers),

Thanks for the suggestion – I installed gllamm, and am trying to run it.
I’m running a very basic “trial” model before I do my main runs, but It’s
taking FOREVER to iterate.  The syntax I’m using is:

****identifying data as panel data*****
Tsset c_id index1

***basic model***
Gllamm [outcome] [variable], i(childid schoolid) pweight(we)

Is there a faster way to run weighted, clustered, random-effects regressions
using panel data?  This very basic model has taken three hours in gllamm,
and is still iterating.  I’m not convinced it will converge before I finish
graduate school (eek!).

Thanks for the help!

--Melissa 


On 12/20/10 1:16 PM, "Stas Kolenikov" <skolenik@gmail.com> wrote:

> You can estimate these models with all the features you need using
> -gllamm-. Type
> 
> ssc install gllamm
> 
> to download it; look at the website (http://www.gllamm.org) for
> specific examples. It wouldn't be too difficult in your case, although
> generally -gllamm- syntax can be challenging.
> 
> On Mon, Dec 20, 2010 at 9:22 AM, Castle, Melissa H <mhc14@pitt.edu> wrote:
>> Hi,
>> 
>> This is my first time posting to the list ­ Iąve checked out the FAQs and
>> archives, and haven't been able to figure out the answer to my question.
>> Sorry if itąs basic ­ Iąm learning.
>> 
>> So, Iąm trying to figure out how to run weighted, clustered regressions that
>> vary randomly across children (random effects regressions) using panel data.
>> Is it possible to do this using STATA?  Or do I really need to be doing this
>> kind of modeling in HLM?
>> 
>> ***A little bit about my data***
>> * Iąm using the ECLS-K, a large, longitudinal, nationally representative
>> dataset (n=21,000).  I use a weight to ensure that the data remains
>> nationally representative once accounting for attrition from Kindergarten
>> through 3rd grade.
>> * Iąm also analyzing children who are clustered in schools, so need to
>> account for this nesting.
>> 
>> More specifically, here is what will and will not run:
>> 
>> ***THESE MODELS WILL NOT RUN***
>> xtreg [outcome] [varnames] [iweight=c245cw0], vce(cluster schoolid) re
>> i(c_id)
>> ERROR MESSAGE: Weights not valid with re command
>> 
>> xtreg [outcome] [varnames] [iweight=c245cw0], vce(cluster schoolid) mle
>> i(c_id)
>> ERROR MESSAGE: vce command CLUSTER not valid with mle command ­ only oim,
>> bootstrap, & jackknife
>> 
>> ***THESE MODELS WILL RUN***
>> xtreg [outcome] [varnames] [iweight=c245cw0], mle i(c_id)
>> Xtreg [outcome] [varnames], vce(cluster schoolid) re i(c_id)
>> 
>> Do any of you have a suggestion?
>> 
>> Thanks so much,
>> 
>> Melissa
>> 
>> -------
>> Melissa H. Castle
>> Doctoral Student, Clinical Psychology
>> University of Pittsburgh
>> 
>> 
>> *
>> *   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/
>> 
> 
> 
> 
> --
> Stas Kolenikov, also found at http://stas.kolenikov.name
> Small print: I use this email account for mailing lists only.
> 
> *
> *   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/


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