R Programming Part 14 Covariance in Regressions

The calculations for a general linear model use ordinary least squares will fail if covariance between two variables is 100% (i.e. these two variables are largely the same thing). Also covariance will increase the error in the model that you produce. Luckily R comes up with a few handy functions to identify these pesky covariant variables. You can use these functions to look for improving your model or for forming hypotheses because really knowing two values are roughly the same is in itself an interesting piece of knowledge.