Beyond Type II error

A Type III error is a good solution to the wrong problem. A Type IV error is a wrong solution to the wrong problem.



Statistician’s occupational hazard

An occupational hazard to which we statisticians are exposed occurs in the context of a social occasion, perhaps a dinner party. I am, let us say, seated next to a charming lady whom I have just met, and, as an initial conversational ice-breaking, she turns to me with a winning smile and says: “Now tell me what is it you do?” We must tell the truth, of course, so I reply that I am a statistician. That usually ruins a fine conversation, for in 8.6 cases out of 10 the lady’s smile disappears, she turns to my rival on her other side, and I attack the fried chicken in lonely, misunderstood dignity.

William Kruskal