There's also temporal sample size: you can infer that A causes B if B happens very soon after A, and nothing else around that time might cause B. For example,

There's also temporal sample size: you can infer that A causes B if B happens very soon after A, and nothing else around that time might cause B. For example, if A is exposure to a chemical and B is an allergic reaction. Edit: A great example is that stomach ulcer guy.