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: question about interaction effects with biprobit


From   "Ed Levitas" <levitas@uwm.edu>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: question about interaction effects with biprobit
Date   Tue, 23 Feb 2010 15:32:21 -0600

Dear Statalisters,

Using a biprobit in a partial observbility model, I'm trying to estimate
the marginal effect of an interaction, its standard error and the
resulting significance level.  I've followed a FAQ that describes how to
do this in the standard logit and probit cases
(http://www.stata.com/support/faqs/stat/mfx_interact.html) but am not
sure if my code is correct for the biprobit case.  Regardless of the
equation in which I place the interaction, I obtain identical marginal
effects and se's.  This makes me question my coding.  Could any tell me
if I am estimating the effects correctly?  If not, can anyone point to
my errors?

I reproduced the relevant code below.
Please note that I delimit statements using a semicolon.
My dependent variables, dv and dv2, are equal across all cases since
this is a partial observability model.
The first main effect, repeatall, is a dummy coded 0,1.
The second main effect, indmkt is continuous.

gen irepind = repeatall*indmkt;

xi:biprobit (dv   patentee1 repeatall  indmkt )
            (dv2  patentee2  repeatall  indmkt irepind),
		 tolerance(1e-3) nonrtolerance partial  ;
		 

quietly summarize repeatall if e(sample);
local meanp1 = r(mean);
quietly summarize indmkt if e(sample);
local meancm1 = r(mean);
local xb _b[repeatall]*`meanp1' + _b[indmkt]*`meancm1' +
_b[irepind]*`meanp1'*`meancm1' + _b[_cons];
predictnl dydc = normalden(`xb')*(_b[indmkt] +_b[irepind]*`meanp1') in
1, se(set);
local xb1 _b[indmkt]*`meancm1' + _b[repeatall]*1+
_b[irepind]*1*`meancm1'*1+ _b[_cons];
local xb0 _b[indmkt]*`meancm1' + _b[repeatall]*0+
_b[irepind]*0*`meancm1'*0+ _b[_cons];
predictnl dydd = normal(`xb1') - normal(`xb0') in 1, se(sed);
predictnl dyddt =
normalden(`xb1')*(_b[indmkt]+_b[irepind])-normalden(`xb0')*_b[indmkt] in
1,se(sedt) p(sig);
list dyddt sedt sig in 1;			 

thanks in advance for your time and consideration
Ed
****************************************
Edward Levitas, PhD
Associate Professor
Sheldon B. Lubar School of Business 
University of Wisconsin-Milwaukee
3202 N. Maryland Ave.
Milwaukee, WI  53211
ph: (414) 229-6825
fx: (414) 229-6957
http://www4.uwm.edu/business/faculty/busfaculty/levitas.cfm



*
*   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