Standard errors after matching

[Callam1]

Hi
I am using a linear regression to evaluate the impact of a treatment effect after matching (psmatch) with replacement and K:1 (k>1). What is the easier way to obtain robust standard errors which account for the fact that control observations are being used multiple times? Sample>10k.
Thank you

[Ksharp]

There is PROC ROBUSTREG .
But I am doubt that if it was what you are looking for.

[Callam1]

It would be interesting to know if it has been used to estimate the treatment effect after matching with replacement and k:1.

[Ksharp]

That is not really .
It don't care whether the data is "matching with replacement and k:1." or not,
It is just consider the high leverage point(for X variables) and outliers point(for Y variable),
and get rid of these point to fit a regression model (a.k.a robust mean or std ).

Anyway, @StatDave @Rick_SAS maybe could give you the answer.

[Rick_SAS]

I do not understand the question, so I cannot add to this discussion.

[Callam1]

The question is how to obtain robust standard errors when using a linear regression to estimate the impact of the treatment effect after matching with replacement and k=n, distance=pscore