Multicollinearity - Explained Simply (part 1)

I describe what multicolinearity is, why it is a problem, how it can be measured, and what one can do about it. I also give guidelines for interpreting levels of tolerance and the variance inflation factor.

For more information and references, check out:

[nuttyfruitloaf]

Your videos are so helpful! You make stats less scary. Thank you!

[Value_Pilgrim]

Wow! Understanding in 2 minutes. Thanks!

[escapetothemoon]

2023 lol this video is still worth to watch

[howstats]

With just two independent variables, yes, but if you have more than two independent variables, you need to consult the tolerance levels; they need to be higher than .10

[Mista]

I'm probably the only one here because I'm a writer and I'm trying to use this term in lyrics I'm writing.
Please tell me if this even makes sense with the way I'm using it.


Pico I need you I pray that you're here with me
I cannot overcome this collinearity


Pico is a friend of mine that recently took his life and when trying to rhyme
"Pico I need you I pray that you're here with me"
I couldn't think of a way to rhyme the end of the next line so I got on a rhyming
website and the term Multicollinearity or collinearity rhymes perfectly with the
line before it so I was trying to find a way I could possibly use it that actually
makes sense and doesn't make the smart listeners think "wtf is he talking about?"


lmao

[Std]

What should be the correlation threshold value based on which we determine the highly collinear variables?

[emanweri]

Wonderful explanation, thank you very!!!

[adameu]

so in a nut shell if i had y= family income, and had some constant value, then beta 1 = husband income and beta 2= years of education of husband. you have multi collinearity because your basically double counting the same effect because husband income is strongly correlated with husband's years of education? so to fix you either have to combine the variables or drop one? thanks in advance for any help people :)

[FunTopia-gx]

i just wanna slap that person so hard who introduced Econometrics subject.... God it is so difficult... i am cryingggg :(

[mihaicarnuta]

So, if there is a significant correlation between two or more IDs, but the r is less than .9 or .8, we shouldn't worry about it?
Thank you!

[eteatestspreparation]

Plz tell me about Urdu econometrics videos