If we were really not allowed to say anything about [a hypothesis], significance testing would be completely useless, but in fact it is only mostly useless.
- Many rules of statistics are wrong - Allen Downey