Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


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

st: Heteroskedastic Probit Model


From   Mustafa Brahim <datotanseri@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   st: Heteroskedastic Probit Model
Date   Thu, 22 Apr 2010 16:51:04 +0800

Dear Stata users,


I run this model:

hetprob crisis m2mult m2reser usrealra creditgd reserves rerdev bfl
privclai shordebt shortdeb, het( m2mult  m2reser usrealra creditgd
reserves rerdev bfl privclai shordebt shortdeb )

I included all variables in the het ( ) option because I am not sure
which one may be causing the problem of heteroskedasticity. And this
is the result:
---------------------------------------------------------------------------------------------------------------------
Heteroskedastic probit model      Number of obs     =        833
                                               Zero outcomes     =        708
                                               Nonzero outcomes  =      125

                                               Wald chi2(10)     =       5.73
Log likelihood = -144.4906           Prob > chi2       =     0.8374

-----------------------------------------------------------------------------------------------------------
               |      Coef.         Std. Err.       z       P>|z|
[95% Conf. Interval]
-------------+-------------------------------------------------------------------------------------------------
crisis          |
     m2mult |  -.2181513    .2241422    -0.97    0.330    -.6574619    .2211593
    m2reser |  -.2453235    .5018176    -0.49    0.625    -1.228868    .7382209
   usrealra  |  -2.490834    1.276029    -1.95    0.051    -4.991805
  .0101379
   creditgd  |  -.5277773    .4011932    -1.32    0.188    -1.314102
   .2585468
   reserves  |  -1.951593    1.126167    -1.73    0.083     -4.15884
   .2556536
     rerdev   |  -.8494946    4985961    -1.70     0.088    -1.826725
    .1277359
        bfl      |   .0908733    .1461244     0.62   0.534
-.1955254    .3772719
   privclai    |   2.569919    1.285969     2.00    0.046     .0494657
    5.090372
   shordebt |    -1.7086     .8392513    -2.04     0.042    -3.353502
   -.0636975
   shortdeb |   1.312913    .6240398     2.10    0.035      .089817
   2.536008
      _cons  |   .8284442    .6997099     1.18    0.236     -.542962
     2.19985
-------------+-----------------------------------------------------------------------------------------------
lnsigma2     |
     m2mult |   .0798557    .3237996     0.25    0.805    -.5547798    .7144913
    m2reser |  -.9097904    .6999602    -1.30    0.194    -2.281687    .4621064
   usrealra  |   1.093203    .3208028     3.41    0.001     .4644409
  1.721965
   creditgd  |   .4358112    .6106535     0.71    0.475    -.7610477     1.63267
   reserves |   .6430296    .4342553     1.48    0.139    -.2080951
  1.494154
     rerdev   |  -.9071082    .3128004    -2.90   0.004    -1.520186
    -.2940307
        bfl     |   .5611073    .3041175     1.85    0.065    -.034952
      1.157167
   privclai    |  -1.616153    .6099832    -2.65    0.008    -2.811698
   -.4206083
   shordebt |   1.475549    .6217962     2.37    0.018     .2568507     2.694247
   shortdeb |  -2.131526    .5119575    -4.16    0.000    -3.134945    -1.128108
---------------------------------------------------------------------------------------------------------------
Likelihood-ratio test of lnsigma2=0: chi2(10) =    40.65   Prob > chi2 = 0.0000


I have a couple of questions:

1) I understand that runing hetprob procedure adjusts the estimates
and standard errors and therefore corrects for heteroskedasticity? is
this correct? can I conclude that all variables with Prob

2) if that's the case which table I shouls refer to? the first or the
second which shows lnsigma2?

3) since the Prob > chi2 = 0.0000   is less than 0.005, it shows that
there is a problem of heteroskedasticity. Can we tell from the second
tablewhich are the variables causing the problem?


Appreciate your comments and help,


Ismail
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   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   |   Site index