Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • PSM (ATET and ATE)

    Hello all,

    I will please like an answer to this question. When calculating Average Treatment Effect (ATE) and Average Treatment Effect on the Treated (ATET) I believe results should show different number of observation for each estimate as one is on the general population (ATE) and the other is on the treated (ATET). I however find that both estimates show the same number of observation. How do I get the number of observation for estimates on the treated (ATET)? An example is shown in page 139 in the Stata manual. Pls follow this link. https://www.stata.com/manuals13/te.pdf.
    Many thanks.





  • #2
    You should probably show your commands and the output of these commands. Or is your question just about this example from the Stata manual?

    Comment


    • #3
      The example in page 139 of the manual shows the exact problem in my analysis. Same number of observations are reported for ATE and ATET. I want to know how to get the number of observations for the treated population.

      Comment


      • #4
        I still don't understand what you mean.
        I want to know how to get the number of observations for the treated population.
        Do you want to know how the number of observations is calculated or how to get the number of observations like it is displayed in the output?
        When you look at the manual then you see at page 140 that the number of observations for treatment level j is stored in the scalar "e(n0)" and "e(n1)" which you can access by typing ereturn list.

        Comment


        • #5
          Thanks a lot for your reply. Just to confirm, since the ATET uses the treated population can i put just that number of observation (324) in my thesis for the ATET result and the overall number of population (1134) for the ATE result.

          Please see the result i received from the ereturn command

          scalars:
          e(k_levels) = 2
          e(k_robust) = 2
          e(k_nnmax) = 17
          e(k_nnmin) = 1
          e(k_nneighbor) = 1
          e(control) = 0
          e(treated) = 1
          e(caliper) = .
          e(n1) = 324
          e(n0) = 810
          e(N) = 1134

          Comment


          • #6
            I think that you are correct.

            Comment

            Working...
            X