Hawks' First Axiom of Science

less than 1 minute read

OK, I never thought I would start a list like this, but I have been irritated enough, already! So here it is:

Axiom I

The scatterplot conjecture

An empirical paper with fewer than 100 observations that does not include a scatterplot is to be seriously doubted.

Some selected papers may overcome this presumption of doubt, but scatterplot-avoidance is almost always a strategy to (a) cover up significant outliers, or (b) keep data secret so that the statistics cannot be replicated. Since these motives are enemies of science, the lack of a scatterplot (or even better, the data themselves) must inspire serious doubt.