Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st: something wrong with how I define my biprobit ? or probit? or a bug somewhere
From
Nirina F <[email protected]>
To
[email protected]
Subject
st: something wrong with how I define my biprobit ? or probit? or a bug somewhere
Date
Sun, 28 Mar 2010 09:01:39 -0400
Hello Statalisters,
I would like to estimate the following where y1 and y2 are both
dichotomous variables and endogenous.
y1* = alpha1 + betaX + betay2* + e1 (eq. 1)
y2* = alpha2 + betaX + e2 (eq. 2)
It is true that when rho is zero in a biprobit model of two equations,
then that means that cov(e1,e2) = 0 and that y2 is exogenous so that
one can just estimate equation 1 with simple probit.
Now when I try to restrict rho=0 with my biprobit and then run also
probit with the data and example below, it is wierd that I do not get
the same coefficient on variable private. With biprobit it is: -0.0180
and with probit, it is -0.177
#delimit ;
webuse school;
local athrho=atanh(0);
constraint 1[athrho]_cons=`athrho';
biprobit (private = loginc logeduc years) (vote = private loginc
logeduc years), constraint(1) ;
mfx compute;
outreg2 using schoolStata, mfx e(all);
Now with probit
#delimit ;
probit vote private loginc logeduc years;
mfx compute;
outreg2 using schoolStata, mfx e(all);
Could you please point out if there is something wrong in my
definition or this could be an indication that there is a bug with
biprobit?
Thank you very much,
Nirina
ps:
which biprobit
*! version 2.10.1 20sep2007
which probit
*! version 1.1.0 26feb2007
*
* 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/