Murali Kuchibhotla <[email protected]>:
I think you need to first think about what the relevant treatment
effect is, i.e. what would you compute assuming you observed all L
outcomes, which should tell you how to proceed with matching. If
there is a baseline treatment "control" then you want to match the
other L-1 types to that group and estimate L-1 different treatment
effects (maybe). If there is one treatment group and L-1 "control"
groups, see http://www.biostat.jhsph.edu/~estuart/RubinStuartJSM2005.pdf
On Tue, Oct 28, 2008 at 10:21 AM, Murali Kuchibhotla
<[email protected]> wrote:
> Hello,
> I was wondering if anyone is aware of the existance of a Stata module
> out there that can implement propensity score matching in the presence of
> multiple treatments, i.e., where there are L>2 outcomes{Y(1),Y(2)..,Y(L)} and
> only component of which can be observed for each individual, leaving the other
> L-1 outcomes as counterfactuals.
*
* 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/
