Re: st: Computation of ATT in logistic regression

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

--- On Mon, 22/11/10, Astrid Kiil wrote:
> Does anybody know how to compute the
> average treatment effect for the treated (ATT) of an
> explanatory variables included in a logistic regression in
> Stata? 
> 
> (I want to do so in order to obtain a regression estimate
> that is comparable to the ATT obtained by propensity score
> matching, performed by e.g. psmatch2 or nnmatch).

That is a surprisingly difficult problem, in essence the -logit-
model is a model that thinks of effects in terms of ratios, while
the ATT is typically thought of in terms of differences. An 
effect that is constant in terms of ratios cannot be constant in
terms of differences (except when the ratio is 1 and the difference
is 0, i.e. when there is no effect). So any solution that might 
exist would boil down to fitting a linear model on top of your
non-linear model. In that case, I would just skip the intermediate
nonlinear/logit model, and go directly for the linear model/
-psmatch2-.

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/