RE: st: Test of marginal effects in two samples

["Kleinjans, Kristin" <kkleinjans@exchange.fullerton.edu>]

Dear Maarten:
Thanks for your help. 
Just to make sure I understand exactly what you mean, is this what you propose that I do:

constraint 1 [ident=0]: sample
mlogit work ragender##hcplt_1 edu2##sample i.age##sample ident, cluster(hhidpn) constraints(1)

margins, predict(outcome(1)) dydx(edu2) over(sample)

where sample is the sample indicator, and ident is the variable that is only present if sample==0.

Thanks,
Kristin

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of maarten buis
Sent: Monday, May 09, 2011 12:53 AM
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: Test of marginal effects in two samples

On Mon, May 9, 2011 at 2:32 AM, Kleinjans, Kristin
<kkleinjans@exchange.fullerton.edu> wrote:
> I conduct multinomial logit regressions on two different samples using the same outcome
> variables and (mostly) the same independent variables. I would like to test whether the
> marginal effects  in the two samples are statistically different.

The typical trick for that is to estimated them all in one model and
add interactions of your group identifier with all your explanatory
variables. You can use -constraint-s to remove some explanatory
variables from some of your samples, i.e. set the effect of those
variables in those samples equal to 0.

Hope this helps,
Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany


http://www.maartenbuis.nl
--------------------------

*
*   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/

*
*   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/