Estimation of ATT for a naive regression estimator

[Kate Hampstead]

Hello,
I want to estimate an average treatment effect on the treated (ATT) for a naive regression with 3 covariates. The only solution on the internet I have found was like this:

teffects ra (x1 x2 x3) (treatment), atet

However, I am not entirely sure if it is correct. I know there are some ATT (att*) commands for nearest neighbour matching (attnd) or radius matching (attr), kernel matching (attk) but I am not sure if there is any like this using att* for my problem?

[Jeff Wooldridge]

In Stata, "atet" means ATT. So that is the correct command except that you need to specify the y variable before x1.

The teffects command supports various other estimators -- propensity score weighting, matching, PS matching, doubly robust estimators. The "atet" qualifier gives the ATT in each case.