Statalist


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

Re: st: mim: xtmixed for unconditional models


From   Nailing Xia <nailingxia@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: mim: xtmixed for unconditional models
Date   Thu, 23 Apr 2009 09:33:06 -0400

Thanks, Maarten. I did what you suggested, and it happens that when
estimating the third imputed dataset, it keeps doing the iterations -
showing "Iteration xx: log restricted-likelihood = -xx.xx (backed up)"
again and again. I guess this is the sign for the model being
unidentified in the third dataset?

However, even if I drop the third imputed dataset, there is another
problem with other imputed datasets. The result only includes
coefficient and standard error for the intercept, and coefficients for
the random components, but no standard errors for the random effects
parameters. It also shows a warning, saying that "convergence not
achieved; estimates are based on iterated EM". Is this also an
identification problem? What can I do about it?

Thanks!



On Thu, Apr 23, 2009 at 3:46 AM, Maarten buis <maartenbuis@yahoo.co.uk> wrote:
>
> Apperently your model is not identified in your third imputed
> dataset. That can happen in complex models like -xtmixed-.
> You can check that by typing:
>
> xtmixed dvar if _mj==3 || schid: || childid:
>
> You can "solve" this by either creating a new set of imputed
> datasets and hope that the model is identified in all new
> datasets, or just drop the third dataset (-drop if _mj==3-).
>
> -- Maarten
>
> -----------------------------------------
> Maarten L. Buis
> Institut fuer Soziologie
> Universitaet Tuebingen
> Wilhelmstrasse 36
> 72074 Tuebingen
> Germany
>
> http://home.fsw.vu.nl/m.buis/
> -----------------------------------------
>
>
> --- On Wed, 22/4/09, Nailing Xia <nailingxia@gmail.com> wrote:
>
>> From: Nailing Xia <nailingxia@gmail.com>
>> Subject: st: mim: xtmixed for unconditional models
>> To: statalist@hsphsun2.harvard.edu
>> Date: Wednesday, 22 April, 2009, 9:55 PM
>> Hi,
>>
>> I am estimating multilevel models with mulple imputed data
>> (using
>> -ice-). I have got results from the full model using
>> command - "mim:
>> xtmixed dvar var1 var2 var3 if [wgt] || schid: ||
>> childid:". However,
>> I have problems getting results when estimating the
>> unconditional
>> model (i.e., no independent variables, only intercept). I
>> have tried
>> two commands - "mim: dvar if [wgt] || schid: ||
>> childid:"  and "mim:
>> dvar one if [wgt] || schid: || childid:" (the variable
>> one equals to 1
>> for all observations). Stata has run for days without
>> giving any
>> estimates (whereas it took 7 hours for Stata to estimate
>> the full
>> model) - it seems to be stuck at estimaing for m=3. I have
>> tried both
>> commands without using -mim-, and both worked fine and gave
>> the same
>> results. Does anyone know why the unconditional model take
>> so long
>> when using -mim-? Is there a faster way to do it?
>>
>> Thank you.
>>
>> Nailing Xia
>> *
>> *   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/



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