All Models are Wrong, Some are Useful

We've implied the limitations and usefulness of models, but it is important enough that it is worth emphasizing. All models are only approximations of reality. They are useful, but they are never perfect.

The statistician George Box is wrote "All models are wrong. Some are useful." Here he expands on what he means:

...there is no need to ask the question "Is the model true?". If "truth" is to be the "whole truth" the answer must be "No". The only question of interest is "Is the model illuminating and useful?" (Box, 1979)

We will return to this point repeatedly as we evaluate the models we use in this course. For any model, we should be able to answer what it is useful for and what its limitations are - and we will challenge you to consider this in the class.

An important corollary of the idea that all models are wrong (i.e., approximations), is that you have to figure out what actually matters to include in the model and what can safely be left out. Here is Box again in a different paper:

Since all models are wrong the scientist must be alert to what is importantly wrong. It is inappropriate to be concerned about mice when there are tigers abroad. (Box, 1976)