Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: model for longitudinal ordinal data


From   James Shaw <shawjw@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: model for longitudinal ordinal data
Date   Fri, 10 Apr 2009 16:39:33 -0500

Since -gllamm- estimates in a mixed-effects framework, I do not
believe you can specify covariance patterns for the data.  To my
knowledge, ordinal data models cannot be fitted by GEE in Stata.  The
distinction between random effects and GEE often makes little
difference.  I would agree with Jason's comment that fitting a simple
ordinal logit or probit model with a robust variance estimator is the
way to go.  Random effects buys you efficiency at the expense of bias
and inconsistencuy if the underlying assumptions are not met.  I tend
to prefer robust, if inefficient estimators.  In any case where the
simple ordinal logit (or probit) model would fail to produce
consistent estimates, random effects would also do so.

--
Jim




On Fri, Apr 10, 2009 at 3:57 PM, Shubhabrata Mukherjee
<joy_stat@yahoo.com> wrote:
> Thanks a lot James.
>
> I researched -gllamm- but I was not sure if one can specify the type of covariance sturucture in it. I have three time points only and wanted to specify cov(uns).
>
> The -ologit- with the vce(cluster id) should work.
>
>  Joey
>
>
>
> ----- Original Message ----
> From: James Shaw <shawjw@gmail.com>
> To: statalist@hsphsun2.harvard.edu
> Sent: Friday, April 10, 2009 12:08:19 PM
> Subject: Re: st: model for longitudinal ordinal data
>
> There are three programs that can be used to fit random effects
> ordinal probit models in Stata:
>
> -reoprob-
> -regoprob-
> -gllamm-
>
> -regoprob- is capable of fitting generalized random effects ordinal
> probit models that do not entail the proportional odds assumption.
> -gllamm- can also fit random effects (and random coefficients) ordinal
> logit models.
>
> Finally, you may consider simply fitting an ordinal logit or probit
> model to your data with a cluster-robust variance estimator using
> -ologit- or -oprobit-, respectively.  In each case, you would need to
> use the "cluster(subject id)" option.
>
> Hope that this information is helpful.
>
> Regards,
>
> Jim
>
>
>
> On Fri, Apr 10, 2009 at 1:56 PM, Shubhabrata Mukherjee
> <joy_stat@yahoo.com> wrote:
>> Is there any model in STATA which fits longitudinal ordinal data?
>>
>>  Thanks,
>>
>>      Joey
>>
>>
>>
>>
>> *
>> *   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/
>>
>
>
>
> --
> James W. Shaw, Ph.D., Pharm.D., M.P.H.
> Assistant Professor
> Department of Pharmacy Administration
> College of Pharmacy
> University of Illinois at Chicago
> 833 South Wood Street, M/C 871, Room 252
> Chicago, IL 60612
> Tel.: 312-355-5666
> Fax: 312-996-0868
> Mobile Tel.: 215-852-3045
>
> *
> *  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/
>



-- 
James W. Shaw, Ph.D., Pharm.D., M.P.H.
Assistant Professor
Department of Pharmacy Administration
College of Pharmacy
University of Illinois at Chicago
833 South Wood Street, M/C 871, Room 252
Chicago, IL 60612
Tel.: 312-355-5666
Fax: 312-996-0868
Mobile Tel.: 215-852-3045

*
*   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   |   What's new   |   Site index