We suspect the presence of multicollinearity in the model when regression coefficients have the opposite sign from what we would normally expect.

Question

Quizplus · Accepted Answer

Multicollinearity can cause regression coefficients to have signs that are opposite from what is expected based on domain knowledge or logical reasoning, as it indicates that predictor variables are highly correlated with each other, distorting their individual effects on the dependent variable.

We Suspect the Presence of Multicollinearity in the Model When