Two Stage Least Squares - an introduction

[@MrThefonzzi]

You may have saved me, academically. Everyone in my course uses your videos, btw ;) Thanks!

[@maximodrake2768]

i know Im asking the wrong place but does someone know a tool to log back into an instagram account? I was dumb lost the login password. I appreciate any tricks you can offer me

[@treybrayden1622]

@Maximo Drake Instablaster :)

[@maximodrake2768]

@Trey Brayden I really appreciate your reply. I got to the site through google and im trying it out now. Seems to take quite some time so I will get back to you later with my results.

[@maximodrake2768]

@Trey Brayden it worked and I now got access to my account again. I am so happy:D Thanks so much, you really help me out !

[@treybrayden1622]

@Maximo Drake Happy to help :)

[@flavianepomuceno7465]

Thank you very much, Ben! You helped me a lot. Very objective and clear. I will love to see more of your videos.

[@rafaelkovashikawa2711]

almost 3 years after graduation here i meet again my old friend ben...

[@remcomeijer1842]

You are the best mr. Lambert!

[@johannes469]

This is incredibly helpful in understanding this topic!

[@mmad29]

Hi, I got a bit confused, could you explain why we include z1 in the first stage regression? Why do we regress x^ also on the other covariates in the initial model and not only on the instruments for it?

[@SpartacanUsuals]

mmad29 Good question. The answer I have for this isn't particularly illuminating, so forgive me. It just so happens in Monte Carlo studies that the performance of the estimator is better if this is done. Hope that helps. Best, Ben

[@mmad29]

Ben Lambert Ok, thanks for your fast reply!

[@88kamikazejoe]

+Ben Lambert It is necessary to include all exogenous variables (2st) also in your first stage regression, to get an unbiased estimator (even if the variance may increase). You assume that E[z3 e] =0, but we don't know that, therefore it is not necessarily a consistent estimator. Stata is doing that alwyays by default.

[@emmq6630]

ben lambett my man!!!! helping me out again and again it dont stop. moscow17 or zone 2? haha good shit

[@AhmedThahir2002]

For the first stage - Can we use a non parametric model? - or should we only use OLS?

[@gabrielwong1991]

Ben will you ever do this video in matrix form? I get the idea now but just cant get my head round in matrix form!

[@MrEnjoyBeats]

Hello, I have a regression equation and I'm supposed to check that there's no causality from the Y to the independent variables (X). I used the lagged variables from the X's as instrument variables in the 2SLS-option in a program called Eviews. I got a negative value for R-squared. How do I interpret that?

[@mrsahmad8406]

How and when two stage least square is preferred over ordinary least squre

[@poojasrivastava1513]

You are best ❤.

[@angelamilton5134]

Still waiting for someone to use this least squares on a problem question

[@MAHANTING]

really thankful!!

[@liketheduck]

I appreciate the high def. But the writing is still blurry when full screen, even with the highest def available.

[@lastua8562]

must be your eyes :)

[@yunpengdeng9270]

what's the definition of over identification

[@GarimaGupta]

Where we can obtain more than one numerical value of the estimates.

[@rauf1001]

im sorry ben x

[@JamesMantis]

wouldnt it be two "stages"? why is it two stage

[@jeffreyanderson5333]

I love you. Will you marry me? just kidding, because I really love your videos. You made my life much easier.