I am wondering how many subsample ways or to get estimates for some sub-categories of the data without sub-sampling.

So far I know two approaches:

1> We can divide the whole sample into two subsamples and run the same main regression equation for these two subsamples and then comparing the coefficients of the variable of interest.

2> We can add the interaction between dummy representing for one subsample and the variable of interest like this.

I am wondering if we have any other way to do subsample?

  • 1
    $\begingroup$ Sub-sampling means dividing sample. Adding interaction term is not sub-sampling, actually the whole advantage of the method is to get estimates for some sub-categories of your data without sub-sampling. $\endgroup$
    – 1muflon1
    Jul 28, 2021 at 9:26
  • $\begingroup$ @1muflon1, yes, I know what you mean, I will edit the question then $\endgroup$ Jul 28, 2021 at 9:34


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.