Chance, Statistics, and Statisticians

Abstract

Statisticians and statistics deal with the effects of chance events on the treatment of empirical data. I am not attempting to add still another definition to the many already in print; Kendall (1948, p. 1) notes that Willcox listed well over a hundred definitions in 1935. I merely hope my description is neither too narrow nor too broad to describe your own activities as statisticians. Some of your other activities may not be well covered because the statistician's work so often penetrates other fields, a feature of statistics I shall emphasize. My aim today is to illuminate the consequences of that description for the practice, for the writing, and for the teaching of statistics. Also I shall relate its consequences to the structure and activities of our Association. I perceive it as both my task and privilege to present here a broad and philosophical view, rather than the kind of narrow, technical exposition we statisticians usually give in our sessions. I want to entertain not only my fellow statisticians but also their husbands-or their wives.