What is multicollinearity?

• A. Low correlation among independent variables
• B. Zero correlation among independent variables
• C. High correlation among independent variables causing unstable coefficient estimates
• D. Correlation between independent and dependent variables

Answer: (C)

Multicollinearity is when independent variables are highly correlated with each other. This makes it hard to isolate each predictor’s unique effect on the outcome because they carry overlapping information. As a result, the estimated coefficients become unstable: standard errors inflate, confidence intervals widen, and small changes in the data can lead to large swings in the estimated effects or even flip their signs. The key issue is the high correlation among the predictors themselves, not how each predictor relates to the dependent variable. If the predictors are only weakly correlated, or if they are simply correlated with the outcome, multicollinearity isn’t the problem. In practice, when multicollinearity shows up, you might drop redundant variables, combine them, or use methods like ridge regression to stabilize the estimates.