(Im)Perfect multicollinearity

Multicollinearity is a phenomenon in which two or more variables in a multiple regression model are highly correlated with each other. It leads to unreliable estimates of the regression coefficients and difficulty in interpreting the effect of the independent variables on the dependent variable. Imperfect multicollinearity occurs when the correlations between the independent variables are not as high but the coefficients still tend to be unreliable. This can happen in models with large numbers of variables and is difficult to detect by looking at the correlation matrix alone.