Peter Flom’s Statistics 101: How to go wrong with the mean