ABSTRACT
Statistics is about taking measurements, of any sort, and using them to create a model which has some predictive power. There would be little sense in setting up a model on data which is misleading, such as all cows have four legs, the kitchen table has four legs, therefore it is a cow. The data must be sensibly applicable. Because we tend to create models of the world around us and apply the data we tend to forget that this is what we do all the time, so we recognise a cow as fitting our idea of a cow and a table our idea of a table. Where our internal models can fall down are situations that are either unique or where the model is inaccurate. An example of this would a child be trying to decide what group of animals a duck-billed platypus belongs to. It is furry, but lays eggs; it has webbed feet and a beak, but suckles its young. Without a framework to put such an animal in, it is difficult to decide what it is.
