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 on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: mfx after zinb


From   Maarten buis <maartenbuis@yahoo.co.uk>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: mfx after zinb
Date   Thu, 22 Apr 2010 05:16:12 -0700 (PDT)

--- On Thu, 22/4/10, Benjamin Engelstaetter wrote:
> I am estimating an Zinb-model and getpositive significant
> marginal effects after using mfx despite both coefficients
> for the appropriate covariate in the inflate and the main
> equation are non-significant. 

This is because they test subtely different null hypotheses:

the coefficient in the main equation can be thought of as a
test that the ratio of expected counts equals 1 (actually
the the difference in log expected counts equals 0, but this
boils down to the same thing, see: <http://www.stata.com/
support/faqs/stat/2deltameth.html>)

The marginal effects test the hypothesis that the difference
in expected counts equals 0.

Both are meaningful representations of the effect of a 
variable: the first can be thought of as representing
the effect in relative terms while the latter can be
thought of as representing the effect in absolute terms.

This distinction can be clearly illustrated when one 
looks at interactions. This is done for a logit model
in <http://www.maartenbuis.nl/wp/interactions.html>.

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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index