Data Science

Gertrude Cox was an influential statistician, working at a time when Statistics was spreading rapidly as a discipline with immense potential. Her comments on Statistics in the late 1950s are just as appropriate today in the broader field of Data Science:

“Statistician” is a mere word denoting a certain aspect in a human being. The fact that you, as an individual, are classified as a “statistician” does not free you from obligations and responsibilities toward other human beings. Some of you may feel that if you work hard and produce three published articles per year, then you are not to be held responsible for the consequences of your work. However, your food, shelter, clothing, physical and even social comforts and pleasures are provided by others.

You have an obligation to clarify the foundations of your techniques and methods for your clients. Our statistical methods should be tailored to the needs of the users, even if this calls for approximate methods. I want you young statisticians not to become men of success but rather to become men of value. Let me give you one quotation from Einstein, “He is considered successful in our day who gets more out of life than he puts in. But a man of value gives more than he receives.”

I do not decry individual effort, but, in considering the challenge of arid lands, we must have teamwork; that teamwork should be between individuals, universities, and research groups in this and other countries.

One thing is certain, we are at the beginning of a new age—an age that will be richer and will offer more and more opportunities for people whose minds are flexible and who are eager to increase their area of awareness.

How big is the challenge and how far do these frontier areas extend? My reply would be like that of the taxi driver when the judge asked him, “How far can you see?” His reply, “The sun.”

Cox, G.M. Statistical frontiers. Journal of the American Statistical Association, 52(277), 1957.