Fred Mosteller  

Born  December 24, 1916 Clarksburg, West Virginia, USA 
Died  July 23, 2006 Arlington, Virginia, USA 
Residence  USA 
Nationality  American 
Fields  Statistician 
Institutions  Harvard University 
Alma mater  Carnegie Institute of Technology Princeton University 
Doctoral advisor  Samuel S. Wilks and John Tukey 
Doctoral students  Persi Diaconis Stephen Fienberg 
Known for  Statistics education 
Charles Frederick Mosteller (December 24, 1916  July 23, 2006, usually known as Frederick Mosteller or Fred) was one of the most eminent statisticians of the 20th century. He was the founding chairman of Harvard's statistics department, from 1957 to 1971, and served as the president of several professional bodies including the Psychometric Society, the American Statistical Association, the Institute of Mathematical Statistics, the American Association for the Advancement of Science, and the International Statistical Institute.
Contents 
Frederick Mosteller was born in Clarksburg, West Virginia, on December 24, 1916 to Helen Kelley Mosteller and William Roy Mosteller. His father was a highway builder. He was raised near Pittsburgh, Pennsylvania and attended Carnegie Institute of Technology (now Carnegie Mellon University). He completed his ScM degree at Carnegie Tech in 1939, and enrolled at Princeton University in 1939 to work on a PhD with statisticians Samuel S. Wilks and John Tukey.
He married Virginia Gilroy, whom he met during college and married in 1941, and had two children: Bill (b. 1947) and Gale (b. 1953). They lived in Belmont and spent summers in Cape Cod.
Mosteller worked in Samuel Wilks's Statistical Research Group during World War II on statistical questions about airborne bombing. He received his Ph.D. in mathematics from Princeton University in 1946.
He was hired by Harvard University's Department of Social Relations in 1946, where he received tenure in 1951 and served as acting chair of 19531954. He founded the Department of Statistics and served as its first chairman from 19571969, 1973, 19751977. He chaired the Department of Biostatistics at the Harvard School of Public Health in 19771981 and later the Department of Health Policy and Management in the 1980s. He also taught courses at Harvard Law School and Harvard's Kennedy School of Government.
He worked with a statistical assistant Cleo Youtz from the 1950s until his departure from Harvard in 2003, and had an administrative assistant. While most faculty were limited to one office, he used four offices in the Harvard statistics department. He was wellknown for being a good writer, insisting on doing up to fifteen drafts of a paper or book chapter before showing it to his colleagues and an additional several drafts before submitting the paper to a journal.
Mosteller retired from classroom teaching in 1987, but continued working and publishing at Harvard through 2003. He then moved to Arlington, Virginia, so as to live closer to his children, on January 3, 2004.
Mosteller wrote over 50 books and over 350 papers, with over 200 coauthors.
Some of his work involved research evaluation and synthesis, especially in medicine and public health,
Mosteller and David Wallace studied the historical problem of who wrote each of the disputed Federalist papers, Madison or Hamilton. In 1964 this paper appeared on the cover of Time magazine, the first time a paper in the field of statistics received such attention. The Federalist Papers study was conducted in order to demonstrate the power of Bayesian data analysis and required a great deal of computational power for that time.
He was an avid fan of the Boston Red Sox and attended the games in the run up to the World Series. He conducted perhaps the first academic investigation of baseball after his favorite team, the Boston Red Sox, lost the 1946 World Series.
Mosteller cared enormously about the teaching of statistics. Since he was an early figure in this field, he was a mentor to many, his positive attitudes towards teaching influenced his many students.
He used several methods of teaching.
Mosteller said that a lecture should always be "specific, general, specific": begin with a particular example which piques student interest, continue with a general point, and end with a particular example illustrating the general point.
Prompted by a seminar by Derek Bok, in the last 2 or 3 minutes of the class, he would ask the students to write down what was the muddiest point in the lecture and what they'd like to know more about.
He practiced every lecture that he gave at least once, in the real circumstance, so that he could be aware of the timing of the lecture, and so he would not be tempted to speak quickly in order to fit in more material. Instead, he would cut out large chunks of the lecture.
Mosteller taught a class in probability and statistics on an educational program Continental Classroom in 196061, supported by the Ford Foundation and broadcast on NBC: 75,000 students took this class for credit at 320 colleges and universities around the country, and 1.2 million watched the lectures on television on 170 stations. The show received such a large number of viewers despite its being broadcast at 6:30 am. Monday, Wednesday, and Friday covered the statistical material, and Tuesday and Thursday were problem sessions.
Mosteller's graduate students included Janellen Huttenlocher, Persi Diaconis and Stephen Fienberg.
Mosteller is said to have accomplished so much during his life in part because of his good work habits and his abilities to work with others.
Despite being involved in an enormous number of projects, his collaborators noted that he stayed current with all of them. Mosteller kept a list of all projects on a single sheet of paper with their current status and reviewed this sheet daily.
He started meetings on time and ended them on time, even if the meeting did not accomplish its desired goal. This provided an incentive for everyone to prepare for the meeting in an effective way, so that the meetings increased in productivity.
He advised his graduate students never to venture an opinion about the policy that they evaluate, but only about the statistical details.
Frederick Mosteller and John W. Tukey: A Conversation moderated by Francis J. Anscombe, Statistical Science Vol. 3, No. 1 (Feb. 1988), pp. 136144.
