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

st: RE: estout after biprobit

From   "Jann Ben" <>
To   <>
Subject   st: RE: estout after biprobit
Date   Fri, 2 Sep 2005 23:37:54 +0200

Friedrich Huebler wrote: 
> I would like to create a table with marginal effects after -biprobit-
> but I do not understand how this can be done with -estout-.

The problem with -mfx- and multiple equation models is that the
returned marginal effects are associated only with variables, but
not with equations per se. Thus, -estout- just prints the effects 
for all occurences of the variables in the table. 

In Friedrichs case, the solution would be to select just one equation 
to be printed. For example:

 . use
 . biprobit private vote years logptax loginc
 . mfx, predict(pmarg1)
 . estimates store private
 . mfx, predict(pmarg2)
 . estimates store vote

 . estout private vote, keep(private:) eqlabels(,none) ///
   cells(b se(par)) margin drop(_cons) style(fixed)

                  private         vote
                     b/se         b/se
years           -.0021034    -.0063654
               (.0045059)   (.0055705)
logptax         -.0188772    -.4866575
                (.117949)   (.2165683)
loginc           .0665599      .376985
               (.0933679)   (.1665747)

Note that, technically, in both columns the first equation 
("private") is printed, but that does not matter here because
both equations contain the same variables.

If the two equations would differ, one could use the following

 . estout private vote, equations(1:2) keep(#1:) eqlabels(,none) ///
   cells(b se(par)) margin drop(_cons) style(fixed)

I hope this helps.

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index