lrtest: negative values

[Christoph Wunder <christoph.wunder@sowi.uni-bamberg.de>]

Hi list,

I am testing for the significance of a random intercept in a two-level model 
using lrtest (after xtmixed). Although I use exactly the same dataset and the 
same set of explanatory variables (i.e., the models differ only in the level-2 
random intercept), the log-likelihood of the unrestricted model is smaller than 
of the restricted one. lrtest calculates a negative value of the test statistic.

How can one interpret the result? Any suggestions?

Best wishes,
Christoph


Here is the output for reference:

=====[The restricted model is:]===============================================

Mixed-effects ML regression                     Number of obs      =     42118
Log likelihood = -80900.819                     Prob > chi2        =    0.0000

[parameter-output omitted]

------------------------------------------------------------------------------
  Random-effects Parameters  |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
level1: Identity             |
                   sd(_cons) |    1.10895   .0127514      1.084237    1.134226
-----------------------------+------------------------------------------------
                sd(Residual) |   1.444866   .0057342      1.433671    1.456149
------------------------------------------------------------------------------


=====[The unrestricted model is:]=============================================

Mixed-effects ML regression                     Number of obs      =     42118
Log likelihood =   -81184.3                     Prob > chi2        =    0.0000

[parameter-output omitted]

------------------------------------------------------------------------------
  Random-effects Parameters  |   Estimate   Std. Err.     [95% Conf. Interval]
-----------------------------+------------------------------------------------
level2: Identity             |
                   sd(_cons) |   .0852371   .0274306      .0453627    .1601617
-----------------------------+------------------------------------------------
level1: Identity             |
                   sd(_cons) |   1.190898   .0123003      1.167033    1.215252
-----------------------------+------------------------------------------------
                sd(Residual) |   1.385043   .0060852      1.373168    1.397021
------------------------------------------------------------------------------


=====[lrtest output is:]======================================================

Likelihood-ratio test                             LR chibar2(01)   =   -566.96
(Assumption: rest nested in unrest)               Prob > chibar2   =    1.0000


*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/