Commentary on Harold Hotelling's Articles on the Teaching of Statistics

摘要

It is close to half a century since the first of these two papers by Hotelling was published. The second paper, published nearly a decade after publication of the first, reexamines the issues raised in the first and amplifies on the organizational and teaching personnel issues. For one who reads (or rereads) these papers, the striking fact is the thoroughness with which Hotelling has analyzed the many issues that could and would arise in the teaching of statistics. In 1940, statistics was not yet thought of as an independent discipline of study. The subject was still very young. There were only a handful of pioneers - Hotelling one among them - doing research that can be classified as theoretical or mathematical statistics as we know it today. Many of the personalities that would come later and leave their impact on the history and development of the subject were not yet to embark or had hardly embarked upon their careers in statistics. It is in this context that Hotelling's discussion and analysis of the issues involved in the teaching of statistics stand out so distinctively for the foresight and clarity exhibited. In analyzing the problems associated with teaching of statistics in universities, Hotelling has viewed the topic as consisting of three inseparably connected issues: teaching, research, and consulting. Many of the problems discussed here are typically still with us, some have been resolved the way he foresaw or recommended, and some would come back in different contexts as the subject itself were to grow. It is in this light that these papers are reexamined.