Question about multicollinearity. The VIF and Tolerance values are exactly the same bt in the acceptable range, but the inclusion of an extra variable doesn't strengthen the r and r2 much. Should I remove a variable? They're closely related.