mixed model with svy commands

[Ming wen-2]

Hi there,

Does anyone know how to fit multilevel logistic models while taking
into acount survey design (for example, using svy commands) in stata
10?

Many thanks,

Ming
*
*   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/

[Jielu Lin]

Are you using xt- commands? I'm not exactly sure but I may have a
similar problem. I can't weight in xtmixed. Xtreg, xtlogit and others
have weight option, but there are restrictions:  weights must be
constant within panel, and there is not adjustment for clustering
other than that within panel.

I have been looking for a way to use svy commands with xt- commands
but haven't had any luck. By far all I know is that GLLAMM is able to
do this but haven't tried because it requires too much computational
power.

On Tue, Aug 10, 2010 at 3:20 PM, Ming wen <[hidden email]> wrote:

>
> Hi there,
>
> Does anyone know how to fit multilevel logistic models while taking
> into acount survey design (for example, using svy commands) in stata
> 10?
>
> Many thanks,
>
> Ming
> *
> *   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/

[Ming wen-2]

Thanks, Jielu. GLLAMM may be able to run mixed models with survey
design accounted for but it's very slow. So I guess we'd have to
switch to SAS or SUDANN to run this type of models. I wish Stata could
figure out a way to run this type of models in the future.

On Wed, Aug 11, 2010 at 9:22 AM, Jielu Lin <[hidden email]> wrote:

> Are you using xt- commands? I'm not exactly sure but I may have a
> similar problem. I can't weight in xtmixed. Xtreg, xtlogit and others
> have weight option, but there are restrictions:  weights must be
> constant within panel, and there is not adjustment for clustering
> other than that within panel.
>
> I have been looking for a way to use svy commands with xt- commands
> but haven't had any luck. By far all I know is that GLLAMM is able to
> do this but haven't tried because it requires too much computational
> power.
>
> On Tue, Aug 10, 2010 at 3:20 PM, Ming wen <[hidden email]> wrote:
>>
>> Hi there,
>>
>> Does anyone know how to fit multilevel logistic models while taking
>> into acount survey design (for example, using svy commands) in stata
>> 10?
>>
>> Many thanks,
>>
>> Ming
>> *
>> *   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/

[Stas Kolenikov]

SUDAAN does not have multilevel binary response models, and SAS PROC
GLIMMIX or NLMIXED does not support appropriate designs. Indeed GLLAMM
is the only package to do the task you need (and it still does not
correct for stratification, so it is still limited).

On Thu, Aug 12, 2010 at 8:20 PM, Ming wen <[hidden email]> wrote:

> Thanks, Jielu. GLLAMM may be able to run mixed models with survey
> design accounted for but it's very slow. So I guess we'd have to
> switch to SAS or SUDANN to run this type of models. I wish Stata could
> figure out a way to run this type of models in the future.
>
> On Wed, Aug 11, 2010 at 9:22 AM, Jielu Lin <[hidden email]> wrote:
>> Are you using xt- commands? I'm not exactly sure but I may have a
>> similar problem. I can't weight in xtmixed. Xtreg, xtlogit and others
>> have weight option, but there are restrictions:  weights must be
>> constant within panel, and there is not adjustment for clustering
>> other than that within panel.
>>
>> I have been looking for a way to use svy commands with xt- commands
>> but haven't had any luck. By far all I know is that GLLAMM is able to
>> do this but haven't tried because it requires too much computational
>> power.
>>
>> On Tue, Aug 10, 2010 at 3:20 PM, Ming wen <[hidden email]> wrote:
>>>
>>> Hi there,
>>>
>>> Does anyone know how to fit multilevel logistic models while taking
>>> into acount survey design (for example, using svy commands) in stata
>>> 10?
>>>
>>> Many thanks,
>>>
>>> Ming
>>> *
>>> *   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/
>

--
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/

[Jielu Lin]

Thank you both for the information. This is very helpful.

Stas--could you explain more about the limitation of GLLAMM? I'm not
sure about why it doesn't correct for stratification. Thanks!

Jielu

On Fri, Aug 13, 2010 at 11:17 AM, Stas Kolenikov <[hidden email]> wrote:

>
> SUDAAN does not have multilevel binary response models, and SAS PROC
> GLIMMIX or NLMIXED does not support appropriate designs. Indeed GLLAMM
> is the only package to do the task you need (and it still does not
> correct for stratification, so it is still limited).
>
> On Thu, Aug 12, 2010 at 8:20 PM, Ming wen <[hidden email]> wrote:
> > Thanks, Jielu. GLLAMM may be able to run mixed models with survey
> > design accounted for but it's very slow. So I guess we'd have to
> > switch to SAS or SUDANN to run this type of models. I wish Stata could
> > figure out a way to run this type of models in the future.
> >
> > On Wed, Aug 11, 2010 at 9:22 AM, Jielu Lin <[hidden email]> wrote:
> >> Are you using xt- commands? I'm not exactly sure but I may have a
> >> similar problem. I can't weight in xtmixed. Xtreg, xtlogit and others
> >> have weight option, but there are restrictions:  weights must be
> >> constant within panel, and there is not adjustment for clustering
> >> other than that within panel.
> >>
> >> I have been looking for a way to use svy commands with xt- commands
> >> but haven't had any luck. By far all I know is that GLLAMM is able to
> >> do this but haven't tried because it requires too much computational
> >> power.
> >>
> >> On Tue, Aug 10, 2010 at 3:20 PM, Ming wen <[hidden email]> wrote:
> >>>
> >>> Hi there,
> >>>
> >>> Does anyone know how to fit multilevel logistic models while taking
> >>> into acount survey design (for example, using svy commands) in stata
> >>> 10?
> >>>
> >>> Many thanks,
> >>>
> >>> Ming
> >>> *
> >>> *   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/
> >
>
>
>
> --
> 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/

[Steven Samuels]

CONTENTS DELETED

The author has deleted this message.

[Michael N. Mitchell]

Yes indeed it does handle mixed models with weights and strata as well.

Michael N. Mitchell
Data Management Using Stata      - http://www.stata.com/bookstore/dmus.html
A Visual Guide to Stata Graphics - http://www.stata.com/bookstore/vgsg.html
Stata tidbit of the week         - http://www.MichaelNormanMitchell.com



On 2010-08-13 2.41 PM, Steve Samuels wrote:

> I believe Mplus also does mixed models for survey designs.
>
> Steve
>
>
> On Fri, Aug 13, 2010 at 11:17 AM, Stas Kolenikov<[hidden email]>  wrote:
>> SUDAAN does not have multilevel binary response models, and SAS PROC
>> GLIMMIX or NLMIXED does not support appropriate designs. Indeed GLLAMM
>> is the only package to do the task you need (and it still does not
>> correct for stratification, so it is still limited).
>>
>> On Thu, Aug 12, 2010 at 8:20 PM, Ming wen<[hidden email]>  wrote:
>>> Thanks, Jielu. GLLAMM may be able to run mixed models with survey
>>> design accounted for but it's very slow. So I guess we'd have to
>>> switch to SAS or SUDANN to run this type of models. I wish Stata could
>>> figure out a way to run this type of models in the future.
>>>
>>> On Wed, Aug 11, 2010 at 9:22 AM, Jielu Lin<[hidden email]>  wrote:
>>>> Are you using xt- commands? I'm not exactly sure but I may have a
>>>> similar problem. I can't weight in xtmixed. Xtreg, xtlogit and others
>>>> have weight option, but there are restrictions:  weights must be
>>>> constant within panel, and there is not adjustment for clustering
>>>> other than that within panel.
>>>>
>>>> I have been looking for a way to use svy commands with xt- commands
>>>> but haven't had any luck. By far all I know is that GLLAMM is able to
>>>> do this but haven't tried because it requires too much computational
>>>> power.
>>>>
>>>> On Tue, Aug 10, 2010 at 3:20 PM, Ming wen<[hidden email]>  wrote:
>>>>>
>>>>> Hi there,
>>>>>
>>>>> Does anyone know how to fit multilevel logistic models while taking
>>>>> into acount survey design (for example, using svy commands) in stata
>>>>> 10?
>>>>>
>>>>> Many thanks,
>>>>>
>>>>> Ming
>>>>> *
>>>>> *   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/
>>>
>>
>>
>>
>> --
>> 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/