Dear statalist users,
I know this is probably a question which would be easily answered if I had
worked with panel data before but, since it is the first time I�m trying
to work with xtprobit, I�m not really sure how to read the following
output (see below). Since the Likelihood-ratio test of rho=0 does not
reject the inexistence of unoberserved individual level heterogeneity,
should I just use probit instead of xtprobit? If not, how to I test for
fixed vs random effects? xtprobit is for random effects only. Should I use
xtlogit, fe and xtlogit, re and then use hausman?
. xtprobit d3 vargap divida varg_primaria, re nolog
Random-effects probit regression Number of obs =
376
Group variable (i): id Number of groups =
14
Random effects u_i ~ Gaussian Obs per group: min =
15
avg =
26.9
max =
34
Wald chi2(3) =
43.77
Log likelihood = -82.971132 Prob > chi2 =
0.0000
------------------------------------------------------------------------------
d3 | Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+----------------------------------------------------------------
vargap | -.1911532 .0424292 -4.51 0.000 -.2743129
-.1079936
divida | .0058737 .0038235 1.54 0.124 -.0016202
.0133677
varg_prima~a | -.4549667 .103455 -4.40 0.000 -.6577347
-.2521986
_cons | -2.205895 .3058103 -7.21 0.000 -2.805272
-1.606518
-------------+----------------------------------------------------------------
/lnsig2u | -14 349.3657 -698.7443
670.7443
-------------+----------------------------------------------------------------
sigma_u | .0009119 .1592902 1.9e-152
4.5e+145
rho | 8.32e-07 .0002905 3.5e-304
1
------------------------------------------------------------------------------
Likelihood-ratio test of rho=0: chibar2(01) = 0.00 Prob >= chibar2 =
1.000
*
* 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/
