Variance Inflation Factor [VIF]

PrepNuggets

LEVEL II

The Variance Inflation Factor (VIF) is a measure of how much the variance of an estimated regression coefficient is increased due to multicollinearity in the model. It is used to identify correlated independent variables in a multiple regression model.

We start by regressing each of the independent variables against the remaining independent variables. The R-squared from the regression is used to calculate the VIF of that variable using this formula:

VIFj = 1 / (1-R2j)

A high R-squared will translate to a high VIF. A VIF higher than 10 indicates severe multicollinearity, A VIF between 5 and 10 would warrant further investigation into whether the variables are correlated.