st: dprobit/nocons/mfx question

["Schaffer, Mark E" <M.E.Schaffer@hw.ac.uk>]

Hi folks.

Does anyone know why -dprobit- doesn't allow -nocons-?

What makes this a curious limitation is that -mfx- following -probit-
has no problem with -nocons-.

An example using the toy auto dataset is below.  The two regressors are
inverse dummies (when one is 1, the other is 0) so -nocons- is
appropriate.

Stata 10.1, -dprobit- version 3.2.0.

--Mark


. sysuse auto
(1978 Automobile Data)

. gen domestic=1-foreign

. gen mpg20=(mpg<20)

. dprobit mpg20 foreign domestic, nocons

Iteration 0:   log likelihood = -51.292891
Iteration 1:   log likelihood = -47.240655
Iteration 2:   log likelihood = -47.216939
Iteration 3:   log likelihood = -47.216936
may not drop constant
r(399);

. probit mpg20 foreign domestic, nocons nolog

Probit regression                                 Number of obs   =
74
                                                  LR chi2(2)      =
.
Log likelihood = -47.216936                       Prob > chi2     =
.

------------------------------------------------------------------------
------
       mpg20 |      Coef.   Std. Err.      z    P>|z|     [95% Conf.
Interval]
-------------+----------------------------------------------------------
------
     foreign |  -.7478586   .2962199    -2.52   0.012    -1.328439
-.1672783
    domestic |   .1940281   .1749975     1.11   0.268    -.1489606
.5370169
------------------------------------------------------------------------
------

. mfx

Marginal effects after probit
      y  = Pr(mpg20) (predict)
         =  .46573629
------------------------------------------------------------------------
------
variable |      dy/dx    Std. Err.     z    P>|z|  [    95% C.I.   ]
X
---------+--------------------------------------------------------------
------
 foreign*|  -.2837959      .09834   -2.89   0.004   -.47654 -.091052
.297297
domestic*|   .0766821      .06979    1.10   0.272  -.060106   .21347
.702703
------------------------------------------------------------------------
------
(*) dy/dx is for discrete change of dummy variable from 0 to 1

. which dprobit
C:\Stata10\ado\base\d\dprobit.ado
*! version 3.2.0  26feb2007


-- 
Heriot-Watt University is a Scottish charity
registered under charity number SC000278.


*
*   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/