Had a multicollinearity problem with my data set, and I considered your advice to remove a variable from my regression model and it worked! Thank you so much, sir!
I have started to upload economic lectures. For Microeconomics: ua-cam.com/play/PLZ6b0WaGAsFmlYc2l53DZmGH-xIEnOTIr.html For Macroeconomics: ua-cam.com/play/PLZ6b0WaGAsFmwHLks6IMN3Yz26587fU5v.html
Videos on autocorrelation have been uploaded. Please find the following links: 1. ua-cam.com/video/TJkJ0xT5NjI/v-deo.html 2. ua-cam.com/video/TJkJ0xT5NjI/v-deo.html
Dear Shahab, thank you for your suggestion. Actually, data file will give you the same result as we have discussed in above lecture. Important thing is that after watching the video, student/viewer can handle this problem in their own dataset or not. this could be done only with your own dataset. Feel free to contact me if you have further query.
sir, please, how can we change the mesure of a data, i mean what are the alternative ? if i have for instance a data about number of business per city.
Dear sir, thanks for your good explanation. Here, I have question. In your video, I observed that you randomly added the values and increased the sample size. What will be the possible effect on estimation.? can we apply said method for ARDL estimation ?
Hi! I generated the descriptive statistic and correlation matrix, and I found a multicollinearity problem. Is it correct that I fix the multicollinearity problem first, test the the unit root and then run the panel data analysis. Thank you! Also, as I found the multicollinearity problem from correlation matrix already. Should I analyse multicollinearity through VIF again? Thank you.
Thank you for asking. 1. It is all right to correct multicollinearity but problem is that wether OLS will be applied or not. If it is so, then it is right. But after Unit root test, it will be confirmed wether OLS is appropriate or any other method. May be ARDL/MG/PMG will be used. 2. VIF is more formal method than correlation matrix.
@@TJAcademyofficial Thank you so much. Can I ask if I found a multicollinearity problem, should I delete the variables which has multicollinearity before using Breusch and Pagan Lagrange Multiplier to test whether the pooled OLS model or the random effect model I will use.
@@Wendy-kr7dr Yes you cam delete that variable if said variable is not a focus variable of your research. Focus variable means, your objective of research and research question do not depend on that variable.
Hi, I followed your step to try and test for Multicollinearity in Eviews but this message appears "Near singular matrix error. Regressors may be perfectly collinear."
@@TJAcademyofficial thank you very much for your response. I have opened them as a Group but the option of Variance Diagnostic under View is not visible :(
thankyou sir. sir, if the model we are assuming is POLS so is it necessary to test for multicollenearity. please guide . and refere some paper on panel data analysis using POLS please if any. it will be great help thankyou. waiting for the quick responses sir.
thank you sir for clearing all my doubts regarding multicollenearity... regards always as my true phd guide...
Had a multicollinearity problem with my data set, and I considered your advice to remove a variable from my regression model and it worked! Thank you so much, sir!
My pleasure 😊
thank you sir for your efforts its very important for us to understand the concept of regression model
So nice of you :)
Thanks a bunch to you sir.
Thank u Sir 👍
Booht khoob sir
GOOD AND VERY EFFICIENT CONTENT.
CAN U PLEASE PROVIDE SAME DATA SET
Thank you Sir for uploading this video... Please also upload your lectures on economics topics :)
I have started to upload economic lectures.
For Microeconomics: ua-cam.com/play/PLZ6b0WaGAsFmlYc2l53DZmGH-xIEnOTIr.html
For Macroeconomics: ua-cam.com/play/PLZ6b0WaGAsFmwHLks6IMN3Yz26587fU5v.html
Thank you sir for uploading video kindly also upload autocorrelation videos
Videos on autocorrelation have been uploaded. Please find the following links:
1. ua-cam.com/video/TJkJ0xT5NjI/v-deo.html
2. ua-cam.com/video/TJkJ0xT5NjI/v-deo.html
um using a student version and i cannot edit numbers in eviews how do change the problem of multi colleanerity
Its better to share data sets as well, to fully grasp the concepts.
Dear Shahab, thank you for your suggestion. Actually, data file will give you the same result as we have discussed in above lecture. Important thing is that after watching the video, student/viewer can handle this problem in their own dataset or not. this could be done only with your own dataset. Feel free to contact me if you have further query.
Well explained
Great and simple. What if we take and use logs of variables? will it remove multicollinearity issue?
sir, would you please explain the decomposition techniques, how and why are they used?
sir, please, how can we change the mesure of a data, i mean what are the alternative ? if i have for instance a data about number of business per city.
hello sir, this VIF option is not available in my eview software can you help me in this
Dear sir, thanks for your good explanation. Here, I have question. In your video, I observed that you randomly added the values and increased the sample size. What will be the possible effect on estimation.? can we apply said method for ARDL estimation ?
Could you please explain what is the difference between fdi and fdir? Is the latter one the first difference or the lagged version of the former one??
Fdir is the growth rate of fdi
When u removed fdi and ran a model again, your durbin statistic was less than r square. It’s a sign of spurious regression
Thank you for your message. Yes it is but in this video the focus is on Multicollinearity only and it's removals. 👍
where file
Sir
R u conducting any workshop about this study
I am not getting any option for VIF for my panel data in eviews.... What could be the reason?
Not available for panel data
@@TJAcademyofficial so how to find out this for panel data?
I mean all diagnostic tests for panel data
Heteroscedasticity
Multicollinearity etc
Sir can we use these method of removal of multicollinearity after estimating ARDL model??
ARDL utilize Koyck model which includes removal of multicollinearity
And how do we do that sir..is there any video of yours regarding this??
Sir I have 4 independent variables but only one variable is high then 10 please guide
Change its measure
Hi! I generated the descriptive statistic and correlation matrix, and I found a multicollinearity problem. Is it correct that I fix the multicollinearity problem first, test the the unit root and then run the panel data analysis. Thank you! Also, as I found the multicollinearity problem from correlation matrix already. Should I analyse multicollinearity through VIF again? Thank you.
Thank you for asking.
1. It is all right to correct multicollinearity but problem is that wether OLS will be applied or not. If it is so, then it is right. But after Unit root test, it will be confirmed wether OLS is appropriate or any other method. May be ARDL/MG/PMG will be used.
2. VIF is more formal method than correlation matrix.
@@TJAcademyofficial Thank you so much. Can I ask if I found a multicollinearity problem, should I delete the variables which has multicollinearity before using Breusch and Pagan Lagrange Multiplier to test whether the pooled OLS model or the random effect model I will use.
@@Wendy-kr7dr Yes you cam delete that variable if said variable is not a focus variable of your research. Focus variable means, your objective of research and research question do not depend on that variable.
@@TJAcademyofficial if the variable is my focus variable, should I just keep it even though it has multicollinearity? Thank you
@@Wendy-kr7dr ARDL model will be better in said case. Because ARDL based coefficients are free from multicollinearity.
when you put additional data you just did it randomly it is not legal
It is just for understanding only. I did not claim that additional data is real data. I just want to make it easy to understand the logic.
sir if all variables are not significant in regression model then what we have to do further?
Thank you for your message. Transform your variable or change the measure.
@@TJAcademyofficial thank you sir
Well explained sir...
Can we check multicolinearity in cross sectional data with this method ?? Kindly reply me as soon as possibly...Thankuu
Thank you for your message. Yes we can check in cross sectional data also
@@TJAcademyofficial Thanks Allot...😇
Hi, I followed your step to try and test for Multicollinearity in Eviews but this message appears "Near singular matrix error. Regressors may be perfectly collinear."
Thank you for your message. Select all variables and open as a group. You will find at least any two variables which are same.
@@TJAcademyofficial thank you very much for your response. I have opened them as a Group but the option of Variance Diagnostic under View is not visible :(
Dear, first you have to run OLS, then go to View - coefficient diagnostics - variance inflation factor.
Near singular matrix is appeared while running OLS when 2 or more INDEPENDENT VARIABLES are identical in this case neither OLS will be run nor VIF
Sir i m working on panel data. can we use multicollenearity test on my data? please revert
Yes
thankyou sir. sir, if the model we are assuming is POLS so is it necessary to test for multicollenearity. please guide . and refere some paper on panel data analysis using POLS please if any. it will be great help thankyou. waiting for the quick responses sir.
Thank you sir 🙏