0

I am using a 2SLS approach in order to estiamte the effect from increased education. I want to show the effects of the first stage graphically, ie I want to show that the instrument affects schooling positively. I have read something about calculating the residuals but I do not really understand it. Anyone out there knowing what I should do?

Stina
  • 1
  • you increase your chances for a positive answer if you clarify where you have read that – Christoph Hanck Mar 27 '15 at 13:27
  • I read about it in another question.

    http://stats.stackexchange.com/questions/107956/iv-first-and-second-stage-graphs-and-interpretations#comment207751_107981

    "You could run a regression of Z1 on X2 and X3 and then plot the residuals of this regression against X1. The residuals from this regression are the variation in Z1 after partialing out the effect of X2 and X3. "

     – Stina Mar 27 '15 at 13:31

0 Answers0