GLS estimation to correct for serial correlation

This video explains how GLS estimators can be used in the event that the functional form of serial correlation is understood, and the parameters are also known. Check out ben-lambert.co... for course materials, and information regarding updates on each of the courses. Quite excitingly (for me at least), I am about to publish a whole series of new videos on Bayesian statistics on youtube. See here for information: ben-lambert.co... Accompanying this series, there will be a book: www.amazon.co....

[@jawadrafiq2137]

Ben your a legend mate. I have econometrics exam tomorrow and your videos helped me a lot with my revision and actually understanding the topics.

[@odda132]

once again, you're a lifesaver!

[@meysamsojoudi3947]

Excellent explanation.

[@chandrimakhastagir2917]

Thanks a lot.Please upload more videos if possible!

[@danx2932]

Can GLS/WLS still answer the question of what's x's impact on y? since we have transformation on x and y already

[@lastua8562]

Good question. It should capture the effect using a very similar intuition. Compare it to a log transformation, it still captures the effect. In this case, it transforms both sides of the equation. Let us know if you found a detailed answer in the meanwhile!

[@shraddhabhatia4203]

I don't get it. Why is the OLS not blue now, after deducting one observation??

[@davidiversen2240]

Hi Ben, love the videos, extremely helpful. I have a question though which is an extension of what you were ending the video on, which is adding back that first variable under Prais Winston. I understand the theory of it, however not the algebra of why you multiply the regression for the first observation by (1-p^2)^(1/2), i understand it is to get errors with the same variance, but why multiply it by this specifically? if you could offer any help that would be great, so and quickly is the variance of the errors calculated in a geometric sum as they are continuously correlated with past errors terms?

[@inkyukang2358]

I though GLS is for correcting heteroscedasticity, not autocorrelation... could you explain a little bit about this?

[@lastua8562]

WLS is useful if we have HS in the population. GLS if both HS and SC.

[@tag_of_frank]

You are king

[@farahpangestuty3929]

really helpful. thanks

[@jurliasentosa3395]

thank you so much

[@simsim6579]

Plz can you do for us an example as vidéo about GLS step by step .. on Excel. Thank you