Which option best describes what the Variance Inflation Factor (VIF) measures?

Which option best describes what the Variance Inflation Factor (VIF) measures?

• A. VIF measures how much the variance of a coefficient is inflated due to multicollinearity
• B. VIF assesses the overall model fit
• C. VIF quantifies residual variance across observations
• D. VIF indicates correlation between predictor and response variable

Answer: (B)

VIF tells you how much the variance of a regression coefficient is inflated because predictor variables are correlated with each other. It’s calculated by regressing a given predictor on all the other predictors and looking at R-squared: VIF = 1/(1 − R²). If that R² is high, indicating strong multicollinearity, the VIF is large, meaning the coefficient estimate becomes less precise (its standard error is larger). This is about the precision of individual coefficient estimates, not about how well the overall model fits, the residual spread, or the basic correlation between a predictor and the response. So the correct way to describe VIF is that it measures the inflation in a coefficient’s variance due to multicollinearity.

VIF tells you how much the variance of a regression coefficient is inflated because predictor variables are correlated with each other. It’s calculated by regressing a given predictor on all the other predictors and looking at R-squared: VIF = 1/(1 − R²). If that R² is high, indicating strong multicollinearity, the VIF is large, meaning the coefficient estimate becomes less precise (its standard error is larger). This is about the precision of individual coefficient estimates, not about how well the overall model fits, the residual spread, or the basic correlation between a predictor and the response. So the correct way to describe VIF is that it measures the inflation in a coefficient’s variance due to multicollinearity.