Thank you so much! We need more professors like you - you spoke without the jargon making the entire concept seem so easy (which it is!). I am sick of professors who make everything sound so complex (just to sound smarter!). Loved this!
i came across the concept multicollinearity few minutes in an article before searching for it on youtube and within 1h30mins, i had an insight of this concept, thank you prof...
Wow.. you explained such complex topic with so much ease and in a simpler form ... I must say as a beginner I was able to understand the concept related to Regression and multicollinearity... Thanks a ton Prof. OBI... Regards
Dear Pat: This is an outstanding explanation that helped me better understand the implications of multicollinearity in machine learning applicaitons. Thank you!
Dear Sir, You have always been at my heart since that meeting in SA, Durban when you challenged me on my paper. I have just learnt so much about the concept of multicollinearity. Thank you.
wowww! this is the greatest explaination I have heard about Multicollinearity! Kindly share your presentations on Heteroskedasticity and Homoskadasticity....
you are just making it look easy, and that is a true teaching. it doesn't matter how much you know as a teacher it matters how you will pass that knowledge to others
Really well explained. Loved it. Understood the topic so well for the first time. Also loved the way you showed it practically on data and on excel instead of just explaining the theory. cleared a lot of doubts. Thank you.
Hi Pat Obi. This video is a very well-presented videos I've ever seen. Thanks so much for posting and sharing your knowledge! Your sharing is such a life saver. Hope to see more from you.
my name is Esther Nwakuwa am in Lisbon studying statistic and information management in nova university (master prissily good to see you teach as a Nigeria whooo
First time around this topic and things are nearly clear for me. Amazing explanation, thanks a ton. QQ, at the end you mention "we have to be careful on multicollinearity" do you have any continuation of this lecture that will show how to circumvent this problem.
Dr. Obi, I really enjoy your videos. I have two continuous variables: rcs(Age, 5) and rcs(GRE_score, 6) that I relaxed the cubic splines on and now I a getting huge VIF values for each of those variables. Does VIF work with variables that have relaxed cubic splines please? Thank you for your important work.
Thanks for your kind comments. Sorry, I'm unfamiliar with the econometric implications or applications of cubic splines and spline interpolation. In any event, collinearity does not nullify the F test. I'd be careful though in interpreting results of the t tests.
HI Pat. Very nicely explained. Helped me a lot. I have One question towards the very end. You mention that dropping the variable is not necessary. Only make sure that you don't try assess the individual impacts of each Xi independently on Y. However the overall regression shows a negative impact of nicotine on Y. so if we don't want to understand individual impact of each X, but just the overall impact of them combined on Y, is having the negative signs in there considered okay (even though it is counter intuitive)?
That's a good question. However the answer is yes because the overall impact is embedded in the entire prediction equation, not on the individual signs or values of the coefficients. But please also bounce this of off other econometricians.
At 18:25 you mentioned how the high VIF of X1 and X2 agrees with the high correlation between X1 and X2. Is it always True that if we see high correlation between two predictors, their VIF will be high too? Also, is the inverse True? (My intuition says high VIF may not show up as high correlation because correlation only considers 2 variables at a time while VIF considers all other predictors)
Prof. I would like to suggest that if you could give the data set, we also practice your exercises and can get more clear picture. I am Tyrone De Alwis from University of Colombo, Sri Lanka
hello, is this the same as if we want to look for endogeneity or simultaneity are those two subject are same as multicollinearity. i watched too many videos trying to figure things out confused about the terminologies used, and i needed to ask you the question and i did please help. another question how can i apply 2sls in SPSS or i don't know if its available in excel ?please help. Thanks
No it's not. Collinearity is a case where the independent variables (X variables) are related to each other and as a result, we're unable to identify their individual impact on the dependent variable, Y. Endogeneity is a case where Y depends on X and X also depends on Y. In such a case, we cannot definitely regard X as an "independent" variable (please ask others too). Sorry, I'm unsure if Excel can handle 2sls or know how it's implemented on SPSS. Try searching on UA-cam. Hope this helps.
Thank you, Prof.Obi, this was a very well-explained and insightful lecture. It helped me to understand the concept of multicollinearity very well.
Thank you so much! We need more professors like you - you spoke without the jargon making the entire concept seem so easy (which it is!). I am sick of professors who make everything sound so complex (just to sound smarter!). Loved this!
So impressive way of explaining with clarity. Thanks Obi.
Dear Sir, this is by far the most well-explained video about multicollinearity. Thank you so much for your help
i came across the concept multicollinearity few minutes in an article before searching for it on youtube and within 1h30mins, i had an insight of this concept, thank you prof...
You're welcome, Selomo
Wow.. you explained such complex topic with so much ease and in a simpler form ... I must say as a beginner I was able to understand the concept related to Regression and multicollinearity... Thanks a ton Prof. OBI... Regards
This is one of the greatest, most well-presented videos I've ever seen. Thanks so much for posting and sharing your knowledge!
Dear Pat: This is an outstanding explanation that helped me better understand the implications of multicollinearity in machine learning applicaitons. Thank you!
Very elaborate explanation. Thanks Prof Pat Obi.
Dear Sir, You have always been at my heart since that meeting in SA, Durban when you challenged me on my paper. I have just learnt so much about the concept of multicollinearity. Thank you.
wowww! this is the greatest explaination I have heard about Multicollinearity! Kindly share your presentations on Heteroskedasticity and Homoskadasticity....
thank you so much for this simple yet comprehensive video
This took the mystery out of multicollinearity for me. Thank you so much for the well presented video
you are just making it look easy, and that is a true teaching. it doesn't matter how much you know as a teacher it matters how you will pass that knowledge to others
Perfect lecture on multicollinearity!!!Thanks Prof Pat Obi
Really well explained. Loved it. Understood the topic so well for the first time. Also loved the way you showed it practically on data and on excel instead of just explaining the theory. cleared a lot of doubts. Thank you.
This was presented with so much calm and ease. Thank you very kindly. You are greatly appreciated Prof. Obi
Great work sir. Was a kind of confused what multicollinearity was all about. But this video gave me an insight into it. I appreciate.
Very perfectly explained Mr. Pat Obi.
Hi Pat, just wanted to say thanks, your lecture was very easy to understand , great stuff, hope to see more from you.
Hi Pat Obi. This video is a very well-presented videos I've ever seen. Thanks so much for posting and sharing your knowledge! Your sharing is such a life saver. Hope to see more from you.
Thank you so much to make this definition so clear, which helps a lot!!!! So many thx to you!!!!
thanks alot. I have been searching for some kind of guidance and this video is a life saver.
than you Dr. Obi, this lecture is very clear and very instructive.
Best video of Multicollinearity... thumbs up
Once again a great video! Super helpful, it makes so much sense now, thank you!
Very nicely explained Professor !!!
Really amazing presentation! Very clear and open from beginning to end.👍🏼
my name is Esther Nwakuwa am in Lisbon studying statistic and information management in nova university (master prissily good to see you teach as a Nigeria whooo
Thank you Professor Realy it is Very Intresting
You're welcome!
Very well explained . Thank you
Thank you so much sir,you explain it in a very clear manner.
thank you Prof Obi, this is a very well explained lesson on multiple collinearity
Thanks for the excellent presentation.
Thank you very much for sharing valuable lecture
Thanks. looking forward to seeing more from you. keep doing videos.
Thank you very much Sir! Highly appreciated!
this was awesome!!! you explained it very well
very clearly explained.Thank U
thanks for the feedback.
hi thank you very much, your voice is so attractive, i enjoyed and i learnt so much from your video!
First time around this topic and things are nearly clear for me. Amazing explanation, thanks a ton. QQ, at the end you mention "we have to be careful on multicollinearity" do you have any continuation of this lecture that will show how to circumvent this problem.
Very clear explanation sir. Thank you so much:)
thank you for the perfect explanation
Great Explanation!
Well explained Sir. Thank you
Life saver .. thanks aloooot !!!!
You explained multicollinearity in 50 seconds whereas my teacher has spent over a week failing to do so.
Well explained!
very much useful. Thanks a ton
It's a god sent material. Long live Man.
Thank you Prof.
thanks well explained
thanks for your help i appreciate it
Thank you
Great explanation
Thank u very much sir
Dr. Obi, I really enjoy your videos. I have two continuous variables: rcs(Age, 5) and rcs(GRE_score, 6) that I relaxed the cubic splines on and now I a getting huge VIF values for each of those variables. Does VIF work with variables that have relaxed cubic splines please? Thank you for your important work.
Thanks for your kind comments. Sorry, I'm unfamiliar with the econometric implications or applications of cubic splines and spline interpolation. In any event, collinearity does not nullify the F test. I'd be careful though in interpreting results of the t tests.
THANK YOU, SEVERAL QUESTIONS WERE ANSWERED IN ONE VIDEO. THERE WILL PROBABLY BE MORE IN THE WAY, BUT YOUR EXPLANATIONS ARE ALLOWING TO MOVE FORWARD.
Awsm❣️
I wish I'd found it sooner
HI Pat. Very nicely explained. Helped me a lot.
I have One question towards the very end. You mention that dropping the variable is not necessary. Only make sure that you don't try assess the individual impacts of each Xi independently on Y. However the overall regression shows a negative impact of nicotine on Y. so if we don't want to understand individual impact of each X, but just the overall impact of them combined on Y, is having the negative signs in there considered okay (even though it is counter intuitive)?
That's a good question. However the answer is yes because the overall impact is embedded in the entire prediction equation, not on the individual signs or values of the coefficients. But please also bounce this of off other econometricians.
Excellent
Great! Thanks 🙏
Amazing sir
Sir, This explanation is marvelous. Pls. Try to develop more tutorials using time series data with Eviews, practical applications.
Thanks. By the way, I do have a number of videos on time series (VEC, ARDL, NARDL, etc.). Just visit my channel to see the playlists.
great video! thanks!!!!!!
Wonderful! I get the concept today. May God increase your wealth of wisdom in simplifying complex and tricky issues.
thank you sir
At 18:25 you mentioned how the high VIF of X1 and X2 agrees with the high correlation between X1 and X2.
Is it always True that if we see high correlation between two predictors, their VIF will be high too?
Also, is the inverse True? (My intuition says high VIF may not show up as high correlation because correlation only considers 2 variables at a time while VIF considers all other predictors)
You're intuition is right. 👍🏽
Thanku sir . How choose a particular model or variable and how we decide our model is best fit .
Your specified model is on your own hypothesis. If it's significant and all the diagnostics are ok, you should be good to go.
great video, thanks...
if possible please share the excel...
Prof. I would like to suggest that if you could give the data set, we also practice your exercises and can get more clear picture. I am Tyrone De Alwis from University of Colombo, Sri Lanka
Ok, I will do so. Thanks.
Sir, do I need to include the cell having X in analyzing the data or just the values under it?
Not sure I understand your question. But you should include labels when running a regression on Excel.
@@PatObi labels i should say. Anyway, thank you for your response sir
hello, is this the same as if we want to look for endogeneity or simultaneity are those two subject are same as multicollinearity.
i watched too many videos trying to figure things out confused about the terminologies used, and i needed to ask you the question and i did please help. another question how can i apply 2sls in SPSS or i don't know if its available in excel ?please help. Thanks
No it's not. Collinearity is a case where the independent variables (X variables) are related to each other and as a result, we're unable to identify their individual impact on the dependent variable, Y. Endogeneity is a case where Y depends on X and X also depends on Y. In such a case, we cannot definitely regard X as an "independent" variable (please ask others too). Sorry, I'm unsure if Excel can handle 2sls or know how it's implemented on SPSS. Try searching on UA-cam. Hope this helps.
that's worth a mint
Thank you Sir