Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


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

Re: st: Nested hypothesis testing with multiply imputed data using xtmelogit


From   Maarten Buis <[email protected]>
To   [email protected]
Subject   Re: st: Nested hypothesis testing with multiply imputed data using xtmelogit
Date   Sat, 30 Nov 2013 16:32:40 +0100

In general Steve is correct. However, the test that the variance of
the random parameter equals 0 is a test on the boundary of the
parameter space: a variance can only be larger then or equal to 0.
Tests tend to perform weirdly in that region. Typically the sampling
distribution of the test-statistic under the null-hypothesis tends to
be a mixture distribution instead of a chi-square. I don't know if a
solution for this type of problem has been developed within the MI
framework, let alone whether such a solution has been implemented in
Stata.

Hope this helps,
Maarten

On Sat, Nov 30, 2013 at 3:32 PM, Steve Samuels <[email protected]> wrote:
>
> See the manual entry for - mi test-.
>
> Steve
>
> Steven J. Samuels
> 18 Cantine's Island
> Saugerties, NY 12477
> USA
>
> On Nov 29, 2013, at 4:03 AM, Edward Lorenz wrote:
>
> Hello,
>
> I am new to Statlist and I am hoping that someone familiar with
> multiple imputation estimation knows whether it is possible to perform
> nested hypothesis testing after using xtmelogit (now meqrlogit) on
> multiply imputed data. I am doing a two-level logistic random
> intercepts model. I would like to test the null hypothesis that the
> estimate for the random parameter is zero. Without multiply imputed
> data I would do this with a likelihood ratio test comparing the empty
> single-level model with the empty two-level model. But I have learnt
> that the log likelihood statistictic is not defined in the case of
> multiply imputed data. Is there another way to do this test after mi
> estimate: meqrlogit.  Any help would be much appreciated.
>
> Edward Lorenz
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/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/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/



-- 
---------------------------------
Maarten L. Buis
WZB
Reichpietschufer 50
10785 Berlin
Germany

http://www.maartenbuis.nl
---------------------------------
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index