8 Outliers

This chapter investigates various outliers in the model. The aim of the chapter is twofold. One is to look for weird things that we can discover about the history of the journals by looking at extreme values in the model. The other is to test the model by looking at the extreme things it says. Usually the most extreme things a model says are the least plausible; they are artifacts of the model not facts about the underlying phenomena. So if the model is not too plausible at the extremes, that should give us some confidence in the other things it says.