In the context of multiple regression, strong correlations among the explanatory variables, which often result in large estimated standard errors and insignificant estimated coefficients. This is essentially a property of the sample and not a problem with the method of estimation, although sometimes ridge regression is used as a remedy. Perfect multicollinearity occurs when some of the explanatory variables are perfectly correlated, i.e. linearly dependent; in this case the redundant variable(s) must be removed from the regression.