I flatter myself that this is the best general thing I have written to explain why you must never, not ever, use a p-value, hypothesis test, "significance", Bayes value, or…