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]

st: Variance components model: why is the level two variance so high


From   liliana <[email protected]>
To   [email protected]
Subject   st: Variance components model: why is the level two variance so high
Date   Fri, 15 Nov 2013 10:48:06 -0800 (PST)

Hello, 

I am estimating a variance components model with xtmelogit because my
independent variable is a dummy. However, I get no odds ratio for the
constant and a really high (meaningless) level 2 variance. Do you know why?
With the estimates table command I get the odds ratio for the intercept,
which is tremendously high: 4456.12.
These are the command and the table stata gives me back. I have added the
laplace option because, otherwise, Stata would have been really slow. 



. xtmelogit y || cluster:, laplace variance or

Refining starting values: 

Iteration 0:   log likelihood = -1069.1034  
Iteration 1:   log likelihood = -1046.5944  
Iteration 2:   log likelihood = -1045.6258  

Performing gradient-based optimization: 

Iteration 0:   log likelihood = -1045.6258  
Iteration 1:   log likelihood = -1016.1863  
Iteration 2:   log likelihood = -993.04989  (not concave)
Iteration 3:   log likelihood = -989.44747  
Iteration 4:   log likelihood = -980.82198  
Iteration 5:   log likelihood = -980.77838  
Iteration 6:   log likelihood = -980.77835  

Mixed-effects logistic regression               Number of obs      =    
12451
Group variable: v001                            Number of groups   =     
3252

                                                Obs per group: min =        
2
                                                               avg =      
3.8
                                                               max =       
17

Integration points =   1                        Wald chi2(0)       =        
.
Log likelihood = -980.77835                     Prob > chi2        =        
.

------------------------------------------------------------------------------
       b5_60 | Odds Ratio   Std. Err.      z    P>|z|     [95% Conf.
Interval]
-------------+----------------------------------------------------------------
------------------------------------------------------------------------------

------------------------------------------------------------------------------
  Random-effects Parameters  |   Estimate   Std. Err.     [95% Conf.
Interval]
-----------------------------+------------------------------------------------
v001: Identity               |
                  var(_cons) |   31.74218   5.709468      22.31162    
45.1588
------------------------------------------------------------------------------
LR test vs. logistic regression: chibar2(01) =   169.51 Prob>=chibar2 =
0.0000

Note: log-likelihood calculations are based on the Laplacian approximation.



Any help would be really appreciated. 

Thanks,
Liliana 




--
View this message in context: http://statalist.1588530.n2.nabble.com/Variance-components-model-why-is-the-level-two-variance-so-high-tp7580453.html
Sent from the Statalist mailing list archive at Nabble.com.
*
*   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