> But you can't just let the data "speak for itself" without an explanation or a theory that interprets the data. If you look at the heart attack data, and you

> But you can't just let the data "speak for itself" without an explanation or a theory that interprets the data. If you look at the heart attack data, and you ignore smoking you end up inventing the mythical Type A personality — but it was data driven. https://en.m.wikipedia.org/wiki/Type_A_and_Type_B_personality_theory https://en.m.wikipedia.org/wiki/Type_A_and_Type_B_personalit...