Thanks for the video! Two days after watching it I find there is an easier way to derive this. Take the expectation of the structural model conditioning on the instrument (this gives you two equations because our instrument is binary) and solve for the Wald estimator : )
Wait, after we get sum z_i*(y_i - y-bar) = n_1(y-bar_1 - y-bar), we can similarly get sum z_i*(x_i - x-bar) = n_1(x-bar_1 - y-bar) - Why don't we just cancel out n_1? In this way we will get beta = y-bar_1 - y-bar / x-bar_1 - x-bar. This must be wrong (because otherwise y-bar_0 = y-bar and x-bar_0 = x-bar), but I don't see why we cannot cancel out n_1 at this stage.
Hi. I was looking a way to do this kind of excercise. I've a question: What book could be recommended to practice exercises that involve demonstrations in Econometrics? Sometimes, I have read Sotck & Watson, Wooldrigde (Intro), but I feel that I need other kind of book to get doing this kind of excercise (like video, for example).
Wooldridge is the typical book for an intro. If you want to look into matrix notation, "Econometric Analysis" by Greene would be a reference (you can probably find an old version for cheap).
Thank you for the help! Exactly what i was looking for
Thank you so much for your video! An absolute pleasure to watch, and extremely clear :)
I'm glad it helped ! :)
Life savior!!!!!!!!!!
Thanks for the video! Two days after watching it I find there is an easier way to derive this. Take the expectation of the structural model conditioning on the instrument (this gives you two equations because our instrument is binary) and solve for the Wald estimator : )
Excuse me, but could you please explain in more detail?
very clear and helpful. Thanks!
Thank you!
GREAT
nice video thx
+Dan08 It is a pleasure :)
Wait, after we get sum z_i*(y_i - y-bar) = n_1(y-bar_1 - y-bar), we can similarly get sum z_i*(x_i - x-bar) = n_1(x-bar_1 - y-bar) - Why don't we just cancel out n_1? In this way we will get beta = y-bar_1 - y-bar / x-bar_1 - x-bar. This must be wrong (because otherwise y-bar_0 = y-bar and x-bar_0 = x-bar), but I don't see why we cannot cancel out n_1 at this stage.
Hi. I was looking a way to do this kind of excercise. I've a question:
What book could be recommended to practice exercises that involve demonstrations in Econometrics? Sometimes, I have read Sotck & Watson, Wooldrigde (Intro), but I feel that I need other kind of book to get doing this kind of excercise (like video, for example).
Wooldridge is the typical book for an intro. If you want to look into matrix notation, "Econometric Analysis" by Greene would be a reference (you can probably find an old version for cheap).
Thank you. do you have any literature for treatmet effect iv wald estimator?
unfortunately not. I am sure there is, I just don't know where.
You gotta stop the clicking sound, damnnnnn that's annoying