dear all
I use Eviews software for panel data regression.
But, I did not find any option to test multicollinearity with panel data regression.
However, for time series data, i am finding the option to test multicollinearity.
Can anybody suggest how to do it in eviews.
Hello!
If you are running FMOLS regression model you can check it from the menu bar of your regression output -> View -> Coefficient diagnostics -> Variance Inflation Factor. If you are running other panel regression models than do as follows: After running your regression model you can save it/give a name from the menu bar of your regression output window (e.g. model1), then write in a command section the following: "model1.varinf".
"model1" is the name of your model (it's up to you what name you will choose) and "varinf" is simply the Variance Inflation Factor (VIF). If your VIF value is less than 5 (from 5 to 10 is fine as well) you don't need to worry about multicollinearity.
I hope it will help.
Hello!
If you are running FMOLS regression model you can check it from the menu bar of your regression output -> View -> Coefficient diagnostics -> Variance Inflation Factor. If you are running other panel regression models than do as follows: After running your regression model you can save it/give a name from the menu bar of your regression output window (e.g. model1), then write in a command section the following: "model1.varinf".
"model1" is the name of your model (it's up to you what name you will choose) and "varinf" is simply the Variance Inflation Factor (VIF). If your VIF value is less than 5 (from 5 to 10 is fine as well) you don't need to worry about multicollinearity.
I hope it will help.
Dear Srikanth Potharla
You can test the multicollinearity in Panel data regression by using VIF test.
Exclusive usage of a software for a problem may not be possible. For instance for an intervention problem there is no software for it. What I do is to model it bit by bit. Programming knowledge is essential for this. So multicollinearity may be measured not directly by software but by programming using the software.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression