I am making an IV regression. When constructing my instrument set I make regression only on the initial observations for all of my id's. Then I obtain the residuals. After having done this I need the instrument set containing the residuals for the rest of my observations for each id. It is not possible for my to make the regression on the other observation because the information used In the regression is not available besides from the initial period observations.
So what I am doing is that I use the residuals from the initial observations regression on the rest of the observations for each id. My question is then what consequences this will have, and if there is literature where this has been done before to support me doing this?
Here is the dataex of som of my data to get a feel of what I have done. As you see I have taking the residuals from first observations on copied them for all other observation with that ID. If you have any question please feel free to ask.
Code:
* Example generated by -dataex-. To install: ssc install dataex clear input double(HHIDPN Waves) float(Mother_age_at_death Dad_age_at_death resid resid1) 3010 1 75 80 52.21634 73.19397 3010 2 . . 52.21634 73.19397 3010 3 . . 52.21634 73.19397 3010 4 . . 52.21634 73.19397 3010 5 . . 52.21634 73.19397 3010 6 . . 52.21634 73.19397 3010 7 . . 52.21634 73.19397 3010 8 . . 52.21634 73.19397 3010 9 . . 52.21634 73.19397 3010 10 . . 52.21634 73.19397 10013010 1 69 53 42.85693 67.67946 10013010 3 . . 42.85693 67.67946 10013010 4 . . 42.85693 67.67946 10013010 5 . . 42.85693 67.67946 10013010 6 . . 42.85693 67.67946 10013010 7 . . 42.85693 67.67946 10013010 8 . . 42.85693 67.67946 10013010 9 . . 42.85693 67.67946 10013010 10 . . 42.85693 67.67946 10038010 1 50 66 44.70567 49.12651 10038010 3 . . 44.70567 49.12651 10038010 4 . . 44.70567 49.12651 10038010 5 . . 44.70567 49.12651 10038010 6 . . 44.70567 49.12651 10038010 7 . . 44.70567 49.12651 10038010 8 . . 44.70567 49.12651 10038010 9 . . 44.70567 49.12651 10038010 10 . . 44.70567 49.12651 10050010 1 0 74 43.94625 22.839033 10050010 3 . . 43.94625 22.839033 10050010 4 . . 43.94625 22.839033 10050010 5 . . 43.94625 22.839033 10050010 6 . . 43.94625 22.839033 10050010 7 . . 43.94625 22.839033 10050010 8 . . 43.94625 22.839033 10050010 9 . . 43.94625 22.839033 10050010 10 . . 43.94625 22.839033 10225010 1 82 45 -17.016924 -32.73683 10225010 2 . . -17.016924 -32.73683 10225010 3 . . -17.016924 -32.73683 10225010 5 . . -17.016924 -32.73683 10225010 6 . . -17.016924 -32.73683 10225010 7 . . -17.016924 -32.73683 10225010 8 . . -17.016924 -32.73683 10225010 9 . . -17.016924 -32.73683 10225010 10 . . -17.016924 -32.73683 10299010 1 0 0 22.37574 -16.85536 10299010 4 . . 22.37574 -16.85536 10299010 5 . . 22.37574 -16.85536 10299010 6 . . 22.37574 -16.85536 10299010 7 . . 22.37574 -16.85536 10299010 8 . . 22.37574 -16.85536 10299010 9 . . 22.37574 -16.85536 10299010 10 . . 22.37574 -16.85536 10325020 1 0 52 -28.805786 -30.038923 10325020 3 . . -28.805786 -30.038923 10325020 4 . . -28.805786 -30.038923 10325020 5 . . -28.805786 -30.038923 10325020 6 . . -28.805786 -30.038923 10372010 1 71 73 -41.62628 -26.359983 10372010 3 . . -41.62628 -26.359983 10372010 4 . . -41.62628 -26.359983 10372010 5 . . -41.62628 -26.359983 10372010 6 . . -41.62628 -26.359983 10372010 7 . . -41.62628 -26.359983 10372010 8 . . -41.62628 -26.359983 10372010 9 . . -41.62628 -26.359983 10372010 10 . . -41.62628 -26.359983 10378010 1 0 70 -37.60967 -25.504025 10378010 4 . . -37.60967 -25.504025 10378010 5 . . -37.60967 -25.504025 10378010 6 . . -37.60967 -25.504025 10378010 7 . . -37.60967 -25.504025 10394010 1 0 66 50.56738 60.01219 10394010 4 . . 50.56738 60.01219 10394010 5 . . 50.56738 60.01219 10394010 6 . . 50.56738 60.01219 10394010 7 . . 50.56738 60.01219 10394010 8 . . 50.56738 60.01219 10394010 9 . . 50.56738 60.01219 10394010 10 . . 50.56738 60.01219 10395010 1 0 70 -3.319559 17.950354 10395010 3 . . -3.319559 17.950354 10395010 4 . . -3.319559 17.950354 10395010 5 . . -3.319559 17.950354 10395010 6 . . -3.319559 17.950354 10395010 7 . . -3.319559 17.950354 10395010 8 . . -3.319559 17.950354 10395010 9 . . -3.319559 17.950354 10395020 1 0 83 19.40187 -18.756739 10395020 3 . . 19.40187 -18.756739 10395020 4 . . 19.40187 -18.756739 10395020 5 . . 19.40187 -18.756739 10395020 7 . . 19.40187 -18.756739 10395020 8 . . 19.40187 -18.756739 10395020 9 . . 19.40187 -18.756739 10395020 10 . . 19.40187 -18.756739 10404010 1 0 0 -31.20806 -11.840034 10404010 2 . . -31.20806 -11.840034 10404010 3 . . -31.20806 -11.840034 end
Thank you very much
Kind regards
Mads
0 Response to Consequences and literature of using one regression residual on all id's
Post a Comment