Re: st: Identifying interaction effect in probit

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

--- On Tue, 30/3/10, Urmi Bhattacharya wrote:
> I am using a probit model and a few of my regressors are
> continuous but most of them are dummy variables(takes
> value 1 or 0).I also have interaction term for the dummy
> variables in the probit model.
> how do i find the estimated change in probability of
> success if associated with the interaction term.i.e if
> the interaction term is [siginificant]


This is actually a surprisingly hard question. There are
two solutions: 

1) You could use -inteff-, which you can download by 
typing in Stata -findit inteff- and follow the links and 
the instructions. This approach is discussed in Edward 
Norton, Hua Wang and Chunrong Ai (2004) "Computing 
interaction effects and standard errors in logit and 
probit models", The Stata Journal, 4(2), p.p. 154--167. 
<http://www.stata-journal.com/article.html?article=st0063>

2) You could estimate your model as a -logit- model and
interpret your effects in terms of ratios of odds rather
than as differences in probabilities, which has the 
advantage of greatly simplifying the interpretation of
interaction terms. I discussed that approach here:
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/