John Tukey

John Wilder Tukey (June 16th 1915 - July 26th 2000) is a Statisticien. Born with New Bedford, Massachusetts, Tukey a control in Chimie of the Université Brown before joining the Université Princeton in 1937 obtained to prepare there its doctorate in Mathématiques.

During the Second world war, Tukey worked with the Bureau of Research on the Control of Fire and collaborated with Samuel Wilks and William Cochran.

After the war, it turned over to Princeton, dividing its time with the Laboratoires Beautiful.

Among several contributions to the Civil society, Tukey was useful on a committee of the American Association of the statistics which produced a report/ratio defying the conclusions of the Rapports Kinsey, statistical Problèmes of the Kinsey Report/ratio on the sexual behavior of the human male .

Taking its retirement in 1985, Tukey died in New Brunswick, New Jersey.

Its statistical interests many and were varied. It is particularly known for its development, with James Cooley of the algorithm of the Transformée of fast Fourier. Among its smaller but as much durable achievements was the terms Logiciel and the Boîte with moustache. In 1977, it introduced the Diagramme of quartiles

It as contributed to the statistical practice and articulated the important distinction between the Analyze of exploratory data and the Analyze of confirmative data, believer as much methodology Statistique placed a too large emphase on the latter. Although he believed in the utility to separate two types of analyzes, he noticed that sometimes, especially in the Natural science, this were problematic and he gave for term to such situations the uncomfortable Science.

HAS D Gordon offered the following summary of the principles of Tukey for the Pratique statistics:

… utility and the limitation of statistical mathematics; importance to have methods of analysis statistics which are robust with the violations of the assumptions underlying their use; the need to pile up the experiment of the behavior of specific methods of analysis with an aim of providing a guide for their use; the importance to allow the possibility that the data influence the choice of the method by which they are analyzed; the need for the statisticians to reject the role of “guard of the proven truth”, and to resist attempts required solutions a-time-for-all and arranged surunifications of a subject; the iterative nature of the data analysis; implications of the capacity growing, the availability and the low price of the data-processing installations; formation of the statisticiens.

External bonds

Memories of John Tukey
one of last the interview of Tukey (1984)

Random links: