Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: xtlogit missing information


From   Marie Olson <marie.olson@wayne.edu>
To   Stata <statalist@hsphsun2.harvard.edu>
Subject   st: xtlogit missing information
Date   Thu, 07 Nov 2002 13:52:42 -0500

Does anyone know why xtlogit would produce coefficients for all
variables in an analysis, but fail to produce standard errors, etc. for
the constant or other variables in the model?  I have included two
examples below:
 xtlogit epc dpc, nolog

bivariate model where dpc is the occurrence of policy change, outcome is
occurrence of violence:
Random-effects logit                            Number of obs
=      1198
Group variable (i) : ccode                      Number of groups
=        30

Random effects u_i ~ Gaussian                   Obs per group: min
=        26
                                                               avg
=      39.9
                                                               max
=        41

                                                Wald chi2(1)       =
21.66
Log likelihood  = -384.59828                    Prob > chi2        =
0.0000

------------------------------------------------------------------------------

         epc |      Coef.   Std. Err.      z    P>|z|     [95% Conf.
Interval]
-------------+----------------------------------------------------------------

         dpc |   1.316804   .2829592     4.65   0.000     .7622141
1.871394
       _cons |  -1.565715          .        .       .
.           .
-------------+----------------------------------------------------------------

    /lnsig2u |   .8472979          .
.           .
-------------+----------------------------------------------------------------

     sigma_u |   1.527525          .
.           .
         rho |   .4149475          .
.           .
------------------------------------------------------------------------------

Likelihood ratio test of rho=0: chibar2(01) =   376.44 Prob >= chibar2 =
0.000

Or a multivariate model, where posdpc is the occurrence of policy
change, epc is the occurrence of violence, regimesc is a regime score
(-10 to 10), and peace is the number of years of peace prior to current
year:
 xtlogit epc posdpc regimesc, nolog

Random-effects logit                            Number of obs
=       863
Group variable (i) : ccode                      Number of groups
=        24

Random effects u_i ~ Gaussian                   Obs per group: min
=         9
                                                               avg
=      36.0
                                                               max
=        41

                                                Wald chi2(2)       =
28.67
Log likelihood  = -260.13778                    Prob > chi2        =
0.0000

------------------------------------------------------------------------------

         epc |      Coef.   Std. Err.      z    P>|z|     [95% Conf.
Interval]
-------------+----------------------------------------------------------------

      posdpc |   1.686509   .4461846     3.78   0.000      .812003
2.561014
    regimesc |   .0650246   .0198366     3.28   0.001     .0261455
.1039037
       _cons |  -1.783155   .2400833    -7.43   0.000    -2.253709
-1.3126
-------------+----------------------------------------------------------------

    /lnsig2u |   1.527534   .1884778                      1.158124
1.896943
-------------+----------------------------------------------------------------

     sigma_u |   2.146346   .2022693                      1.784364
2.581761
         rho |   .5833854   .0139242                      .4918199
.6695379
------------------------------------------------------------------------------

Likelihood ratio test of rho=0: chibar2(01) =   296.92 Prob >= chibar2 =
0.000

Thanks!
Marie

*
*   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/



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