Gauss-Markov assumptions part 1

This video details the first half of the Gauss-Markov assumptions, which are necessary for OLS estimators to be BLUE.

[evap]

Dear Ben Lambert,
It is now exam season at the University of Birmingham. As a second year who hates economics, you are currently my rock and my hope to pass my Econometrics exam. You are a hero, a true legend, a life saviour. If I pass this year, I will dedicate my dissertation to you and if you're ever in Birmingham, let me buy you food.
WRITTEN WITH ALL MY LOVE
and several cans of spar budget energy drink

[wypat]

you make someone who hates econometric now think that it is actually fun, wish there are more great teaching videos on YouTube like yours. thanks a lot!!

[totochandelier]

just making a pause after some 30 videos... you have a wondurful vision of econometrics, everything is so consistent (...). It is just sometimes a bit complicated for a french guy to understand your Manchester accent...but after some vids it sounds like music of econometrics 😉 thank you so much!

[sauravjha]

Thank You so much Ben for your tutorials . These tutorials have motivated me to explore the ideas of econometrics deeply.

[jasonleewkd]

This course is great and a super useful refresher!

[mike]

Very helpful explanation. Distracting pron of linear

[mousquetaire]

Taleb piqued my curiosity in this topic; coming here for a more meaty explanation.

[kejeros]

Why would you make fun of our econometrics lord and savior

[Me-jipn]

2:44 I thought that if there is a multiple that the equation is still linear?

[kottelkannim]

Following the outline of Gauss-Markov theorem proof


I noticed that one suggests an alternative linear estimator.
This alternative linear estimator is then shown to have larger variance of the LS estimator.
This suggested alternative estimator is not only an estimator of a model which is linear in parameters (parameters==LS estimators), but also linear in the dependent variable observations or can be made to be linear in these observations.
In the hope I am not misleading any reader, each of the linearity constraints seem to imply the other .

[guilhermefreire]

If a function is non linear in parameters like y = alpha + beta squared times x + u, couldnt i simply substitute beta squared with another variable say, gamma, and we would get y = alpha + gamma x + u? Wouldn't it be linear then?

[xiao]

Thank you. Very useful for me to prepare my quant interview. Reading takes more time

[ДаниилБолдырев-фц]

Привет из России) its best explanation i ve found

[msam]

Plz do videos on estimability and identifiability

[kottelkannim]

Following the outline of Gauss-Markov theorem proof


I noticed that one suggests an alternative linear estimator.
This alternative linear estimator is then shown to have larger variance of the LS estimator.
This suggested alternative estimator is not only an estimator of a model which is linear in parameters (parameters==LS estimators), but also Linear in the dependent variable.
So linearity of BLUE estimators has both constraints:
A. The model is linear in the estimators.
B. The estimators are linear in the dependent variable observations.

[ppthar]

Just found ur channel which i find pretty useful :D Thnx a lot Ben ^^

[laurabaksa]

Why is random sampling important? (asking more for panel data)

[rojafx]

What is the point of these assumptions, what is their significance?

[doanphamvan]

how did you do it can you share with me, thank you

[victorkamau]

2019 January studying for my CAT tommorow 31st.