To measure multicollinearity, you can examine the correlation structure of the predictor variables. You can also examine the variance inflation factor (VIF), which measures how much the variance of an estimated regression coefficient increases if your predictors are correlated. If the VIF = 1, there is no multicollinearity but if the VIF is > 1, the predictors are correlated. When the VIF is > 5, the regression coefficients are not estimated well. Usually, you should remove highly correlated predictors from the model. Because the predictors supply redundant information, removing them often does not drastically reduce the R2.
If the correlation of a predictor with other predictors is nearly perfect, Minitab displays a message that the term cannot be estimated. The VIF values for terms that cannot be estimated typically exceed one billion.
Multicollinearity does not affect the goodness of fit and the goodness of prediction.
After the investigator subtracts the mean, the investigator repeats the analysis with the new predictor. The VIF values fall below 10. Although the VIF values are still large, the investigator feels more confident in the results with lower multicollinearity.