you don't need to do anything special. the advantage of matching is that
it works for any functional form, including interactions. (intuition:
suppose you exact match on X1 and X2; what would be different if you also
matched exactly on X1*X2?)
Gary
On Sun, 22 Apr 2007, K. Miya Woolfalk wrote:
Hi,
How do you match when your treatment is modeled to interact with one of your
covariates? For instance, if you have a regression of Y on treatment*X1 and
X2, X3, X4, how do we match?
Thanks,
Miya