Home Statistics and Visualization West Coast Stat Views (on Observational Epidemiology and more) A defense of the envelope method