John Wilder Tukey was born in New Bedford, a fishing town on the southern coast of Massachusetts. His father was a high school Latin teacher. His mother was trained as a teacher as well, but was denied employment because her spouse was already employed. So she started a private tutoring business and made John her prize pupil instead of sending him to public school where, she thought, "he would only get lazy." Eventually, John graduated from nearby Brown University with a degree in chemistry, followed by a graduate degree in
mathematics from Princeton. Indeed, he spent most of his life at Princeton, where he founded the university's Statistics Department and taught for decades.
John Tukey soon became one of the most influential statisticians of the late 20th century. Much of his work involved robust analysis, which allows researchers to reach credible conclusions even when their data are flawed. He also introduced new ways of presenting data clearly, including the stem-and-leaf diagrams found in Section 6.13 of the text. And jointly with James Cooley, he developed the Fast Fourier Transform, an algorithm with wide applications in the physical sciences. (It helps astronomers, for example, to determine the spectrum of light coming from a star much more quickly than with other methods.)
But Tukey was a wide-ranging thinker. He was not content with staying within the academy. As he put it, "The best thing about being a statistician is that you get to play in everyone's backyard." Thus, he worked as a researcher for AT&T's Bell Laboratories, became a consultant to the government, corporations, and other organizations, and was a prominent participant in social debates.
His government work included the design of the U-2 spy plane. It inspired (jointly with Lyman Spitzer, Jr.) the Hubble Space Telescope. It led to the recommendation to apply statistical methods to adjust the 1990 U.S. Census in order to count poor urban residents who had been missed. In 1973, President Nixon awarded him the National Medal of Science.
Outside the government, he worked for the Xerox Corporation, Merck and Co., the Educational Testing Service, and the NBC television network (for which he designed polls to predict and analyze elections). And he was a vigorous participant in public debates. Thus, he chaired the committee that linked the use of aerosol spray cans to damage done to the ozone layer. And way back in the 1950s, while working for the National Research Council, he made a splash by criticizing Alfred C. Kinsey's research on sexual behavior. The Kinsey Report had shocked the nation by depicting the country's sexual habits as far more diverse than had been thought. But Tukey thought the work was seriously flawed, using, as it did, convenience samples of people who knew each other (rather than scientific random samples). For years, Kinsey and Tukey clashed. Even at their first meeting, Kinsey asked Tukey to stop singing that Gilbert and Sullivan tune while working. Retorted Tukey: "A random selection of three people would have been better than a group of 300 chosen by Mr. Kinsey."
Although few people know it, Tukey was also an amateur linguist who made significant contributions to the language of modern times. Thus, in the 1940's, he coined the term bit, an abbreviation of "binary digit," which describes the 0's and 1's that have become the basis of computer programs. In the 1950's, decades before the founding of Microsoft, he coined the word software, which, he predicted, would become at least as important as all that computer hardware, then consisting of tubes, transistors, wires, and such. Finally, as readers of text Chapter 15 know, he developed and named the HSD test for determining honestly significant differences between pairs of sample means that are analyzed in ANOVA.
Sources: Adapted from David Leonhardt, "John Tukey, 85, Statistician; Coined the Word 'Software'," The New York Times, July 28, 2000, p. A19.