In my last two posts I talked about Ordinary Least Squares, then extended this discussion to the multiple predictor case and briefly talked about some of the problems that may arise. These problems can include omitted variable bias, heteroskedasticity, non-normality, and multicollinearity. Most of these problems are relatively minor in practice and have easy fixes,...