The Pleasures of Statistics: The Autobiography of Frederick Mosteller
Format: PDF / Kindle (mobi) / ePub
From his unique perspective, renowned statistician and educator Frederick Mosteller describes many of the projects and events in his long career. From humble beginnings in western Pennsylvania to becoming the founding chairman of Harvard University’s Department of Statistics and beyond, he inspired many statisticians, scientists, and students with his unabashed pragmatism, creative thinking, and zest for both learning and teaching. This candid account offers fresh insights into the qualities that made Mosteller a superb teacher, a prolific scholar, a respected leader, and a valued advisor.
A special feature of the book is its chapter-length insider accounts of work on the pre-election polls of 1948, statistical aspects of the Kinsey report on sexual behavior in the human male, mathematical learning theory, authorship of the disputed Federalist papers, safety of anesthetics, and a wide-ranging examination of the Coleman report on equality of educational opportunity.
This volume is a companion to Selected Papers of Frederick Mosteller (Springer, 2006) and A Statistical Model: Frederick Mosteller’s Contributions to Statistics, Science, and Public Policy (Springer-Verlag, 1990).
Frederick Mosteller (1916–2006) was Roger I. Lee Professor of Mathematical Statistics at Harvard University. His manuscript was unfinished at his death and has been updated.
the early ﬁfties. Beyond these ages, KPM did not project. KPM wanted to compare the behavior of people in diﬀerent generations. To do this, they broke their sample about evenly into those under age 33 and those over age 33. Some critics would have preferred that the generations be deﬁned according to birth date rather than age. Table 2.2 gives the accumulative incidence curve for intercourse for U.S. white males who have completed no more than the eighth grade of schooling. The percentages
mention before going on that one of the beneﬁts one gets in serving on an Academy committee is no pay. Except for people genuinely on the staﬀ, experts serve the Academy free of charge. I sometimes argue that this is not as beneﬁcial to the public as it sounds, but let me not derail this discussion. It oﬀered a special beneﬁt here. From the start one could see that the required methods of analysis were going to be arcane from the point of view of non-statisticians. In problems where the outcome
should achieve some minimum standard. The authors (Coleman et al.) of the EEOR wanted to go beyond the ﬁndings of diﬀerences to explain them and possibly to show directions where policy might proﬁtably move. Their approach was through a statistical method called regression analysis. Let me interrupt the narrative to explain the idea. When an outcome like academic performance depends upon several variables, we often believe that, if we could assemble measures of these variables in a manner that
10.1. Virginia at a road job in the late 1930s. Near Christmas of 1938, Dr. Olds told me that the meetings of the Institute of Mathematical Statistics were going to be held at the Book-Cadillac Hotel in Detroit and that he thought I should attend. I said that I didn’t know whether we could aﬀord it, and he, in his bluﬀ way, said that I couldn’t aﬀord not to go. And so my mother and I worked it out. The meeting was very exciting for me. The ﬁrst person I heard speak was William G. Cochran, who
calculations required for the scoring. Players did not know how well they had done, or at least not usually because we were not allowed to talk about what had happened at the various tables for fear of giving information to others who had not yet played the hands. How did she and I learn to play bridge? Although we had often played with Virginia’s parents back in Pittsburgh, neither of us had studied the game. Virginia’s father was a good player and a pithy critic; it is from him we have the