st: Re: Interaction and squared effects in a probit (pa) model

[Maarten buis <maartenbuis@yahoo.co.uk>]

--- On Mon, 11/4/11, Gomez-Mera, Laura wrote:
> We are still not sure how to deal with the marginal effects
> of interacted terms (between two continuous variables) in
> our xtgee model. Norton's 'inteff' seems to work for probit
> models with interaction terms but can it be used when we are
> working with panel data? 

I understand from your previous question that you have a 
special kind of interaction: a squared term. In that case
you typically do not want the marginal effect. A marginal
effect approximates the effect with a linear effect. If you
added a square term you think that there is substantial 
non-linearity, so the linear approximation is not 
appropriate. Having said that, -margins- will give you the
"correct but inappropriate" marginal effect, so there is no
need for -inteff- to interpret squared terms.

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/