Variance Inflation Factors: testing for multicollinearity
HTML-код
- Опубликовано: 22 авг 2024
- This video explains what is meant by 'Variance Inflation Factors', and how these can be used to test for the variables most culpable for causing multicollinearity.
Check out oxbridge-tutor.... for course materials, and information regarding updates on each of the courses. 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....
Best explanation of VIF I've ever heard! Excellent work.
Hi, thanks for your message, and kind words. Glad to hear you liked it! Best, Ben
@@SpartacanUsuals bb
Dear Ben, thereby I have finished all your lectures in Part 1 and 2, and all by taking notes and taking it seriously. I am indebted. Greetings from Turkey
Hi, thanks for your message. If you don't include a variable which is important in determining a dependent variable, then, yes, there is the potential for omitted variable bias. However, in models where two variables are highly collinear, then, by definition, one of the variables will explain a significant proportion of the variance which would have been explained by the other. It won't be perfect, but the hope is that this means practically that only one of the two need be included in the regression. In practice there is always this trade off between endogeneity and multicollinearity. Hope that helps, best, Ben
Thanks very much for all the videos. One question, where can I find a book with many examples of the stats in the second part of your course? Also, I recommend reading The G Factor or Bias in Mental Testing if you want a full run down on IQ research.
Studying for my econometrics exam. Can't remember shit from class due to online teaching (and horrible teacher anyways). Love ya for the compact videos.
PERFECT EXPLANATION. PURE. SIMPLE. EASY.
Thanks for the time you spent on this video. Saludos!!!!
Ben,
Thank you very much for this course. I read the A M Leroi article in the FT that mentioned it just at a time when I wanted to find out more about econometrics, in order to understand a journal article in which panel modelling techniques are applied to data about energy demand in homes (the area in which I'm working). Your course did the trick - I now understand most of the article- and I particularly appreciated the way you explained the mathematical notation and gave practical examples.
Clare
Hey Clare,
I am very much interested in energy demand too. About to start a research master in economics so I wonder - if I would like to work in that area in the future - would you think it is important to research as part of the thesis this topic already or can I easily allow myself to explore other research areas which are unrelated to energy?
Thanks,
Stua
Precise and up to the point. Helped me a lot, especially the part if vif>=5 don't include in regression else include. Thanks a ton.
Finally done with all the videos, Thx. planning to read Wooldridge's book.
Great video! I just needed a quick 5 min on this. Very much appreciated!
Thank you. That really made understand what VIF stands for and whats its purpose.
I think this would be really helpful as one of the ways to reduce number of dimensions/attributes for regression analysis.
Wow, you explained this concept so clearly! Thank you, thank you!!!
Nice explanation, thank you!
Great course, thank you! Now I will try the grad-course!
Just wanted to say thanks for this. Good vid and super intuitive!
Thank you, simple and quick explanation.
Ben, you're GREAT !!
thanks a lot for this very simple and clear explanation
Vey help full thanks❤❤❤
very well explained
Great explanation Ben! Cheers
Always good work as usual !
Best explanation!
Hello there! I was wondering: if you exclude a variable due to multicollinearity, won't you then have an endogeneity bias, considering it is correlated with an independent variable and included in the error term?
Good question!
Excellent video
@BenLambert , I did not understand why do we not compare Rsqaure and compare VIFs instead? is it because Rsquare does not give the degree with which a variable is collinear with other variables and the VIF does this?
Thank you for this explanation. May I ask one question concerning applying VIF for Logistic Regressions? Is it possible to also use VIFs there or should one employ Generalized VIFs for better results because of the use of dummy variables. Thank you and please keep up your awesome work
Your sound level is generally very low. Almost impossible to hear what you are saying :-(
Great video! Thanks! (:
excellent video! thanks :)
Sir, are the playlist videos to be watched in sequence to understand the subject? i.e have you uploaded them in a sequence or i must know which topic is related to other to search for the concepts as i am not an economics student but i am learning statistics,etc so i don't know much about which topics are related to each other.
Thanks a lot!
Hi, great explanation. I have a question: some papers use a threshold of 10 for VIF. What other things do you think are taken into the account to set this threshold at this level. Thanks a lot!
Did you find an answer?
@@lastua8562 Hi, nope.
good
W❤