I followed your instructions and compared it to SAS VIF output. Your calculations were on point. This is great because I do not always have SAS available to me. Thank you very much!
hi, I could quite catch your comments at the very end when commenting on the VIF results - do you mean to say that a VIF over 3 is evidence of multi collinarity?
Great video! Thank you so much! Just a quick question though: in the process of creating the linear model, do you check for multicollinearity with VIF at the start on the imperfect model or do you do it on the final model where several variables have already been omitted based on their P-Values?
Dear Nelisje Thank you for your nice comment Actually in my model, initially I have calculated the VIF value for the individual variables and find the variable with higher to lower VIF value. Then I have made different input combination starting from the variable with higher VIF value. Hope you understand..