A speaker stated in a workshop on applied regression analysis, “In business and the social sciences, some degree of multicollinearity is practically inevitable.” Does this statement apply equally to experimental data?
Ans :No
Because,when the speaker stated in the workshop on applied regression analysis, "In bussiness and the social sciences,some degreeof multicolliartiy is practically inevitable".
It means multicollinearity is not present in all of the data. The multicollinearty is exist only the response variable and some independent variables i.e means among all the independent variables multicollinearity exist only few of variables
For example:
The regression model is
Y=β0+β1X1+β2x2+β3x3+β4x5+β5x5
The multicollinearity is present in between (Y on X1 &X2)
That means it only present in variables X1 & X2 not on all the data
We can detect the multicollinearity by using a correlation matrix, Variation infelction factor .
Get Answers For Free
Most questions answered within 1 hours.