“By the early 1940s, asking about a connection between tobacco and cancer was like asking about an association between sitting and cancer.” If nearly all men smoked, and only some of them developed cancer, then how might one tease apart the statistical link between one and the other?
