At a gathering of the School of Arts and Sciences on Oct. 3, 2023, the next tribute to the life and repair of the late Carl Neracher Morris was unfold upon the everlasting data of the School.
Carl Morris was probably the most achieved statisticians of his technology: his contributions ranged from arithmetic to well being care coverage to sports activities statistics. Morris held appointments on the College of California, Santa Cruz, the RAND Company, the College of Texas at Austin, and the Harvard Departments of Well being Care Coverage and of Statistics. He was a Fellow of the American Statistical Affiliation, the Institute of Mathematical Statistics, and the Royal Statistical Society, and was an elected member of the Worldwide Statistical Institute. He served as the chief editor for Statistical Science and as editor of Idea and Strategies for the Journal of the American Statistical Affiliation.
His dedication as a trainer prolonged past the classroom: he was a valued scholar advisor and a supportive mentor for junior college. Many statisticians relied on his basic work in statistical principle in addition to his beneficiant, knowledgeable counsel. His wry humorousness delighted all who knew him and will rework a boring class or assembly into one thing unexpectedly pleasant. Conversations with Morris have been typically shocking, all the time pleasantly so. He could be describing the baseball pennant race, upcoming tennis matches, some deep connection between chance distributions, or insights on educating.
Morris’s contribution to statistical principle started along with his co-author and lifelong buddy, Brad Efron. Morris and Efron had been undergraduates collectively at Caltech and, later, graduate college students at Stanford. In a seminal collection of papers within the Nineteen Seventies, Morris and Efron confirmed how seemingly disparate concepts may very well be unified into a chic complete and, alongside the best way, demystified a conundrum in statistical principle — Stein’s paradox. Their papers confirmed that prediction may very well be improved utilizing Bayesian fashions with prior distributions estimated from knowledge and that the James-Stein shrinkage estimator, which had shocked the sphere of statistics by exhibiting that predictions may very well be improved by augmenting knowledge with seemingly unrelated observations, was a pure consequence of their strategy. This estimator was proven by their work to be helpful in utilized statistics quite than being merely a theoretical curiosity. The papers are broadly thought-about the muse of recent empirical Bayes, have been broadly used earlier than highly effective computing supplied different approaches, and are nonetheless used at present in some deep studying algorithms. Morris’s eclectic pursuits have been evident even on this theoretical work. Efron and Morris confirmed that end-of-season batting averages for a set of main league baseball gamers have been considerably extra precisely predicted when early season averages for every participant have been changed by the James-Stein estimator utilizing knowledge from all gamers of their dataset. Within the phrases of Andrew Gelman at Morris’s retirement celebration, Efron and Morris “de-paradoxed” Stein’s paradox.
Morris’s long-admired mental generosity was evident in his earliest work. When requested in an interview for the publication CHANCE in regards to the concept of utilizing baseball statistics as an example empirical Bayes, Morris stated that he and his co-author Brad Efron “each admired Charles Stein a lot for his genius and his humanity, we selected this subject, hoping we might honor him by exhibiting that his estimator might work properly with actual knowledge.”
Morris and Efron didn’t merely locate the batting common knowledge: Morris had lengthy been a certifiable baseball statistics fan(atic). In truth, after learning aeronautical engineering at Caltech, Morris figured that, by learning statistics, he would possibly land his dream job — statistician for one of many main league baseball groups. His love of baseball started in his youth. There have been no main league baseball groups west of Kansas Metropolis so, along with watching minor league groups within the Pacific Coast League, Morris adopted baseball via AM radio and morning field scores in newspapers. Many statisticians are baseball followers, some even dabble in baseball statistics, however Morris was not a dilettante. Two of us (David Harrington and Joseph Newhouse) spent a whole bunch of hours as youths enjoying the spinner-based sport All-Star Baseball however by no means tailored disks, as Morris did. He later used multilevel modeling to evaluate whether or not Ty Cobb was really a .400 hitter or whether or not he was solely fortunate in just a few seasons. Sure, he probably was a .400 hitter, concluded Morris. At Harvard, Morris turned the founding college advisor for the Harvard Sports activities Evaluation Collective (HSAC), the place his regular presence and beneficiant counsel impressed college students to start careers in sports activities analytics. As HSAC famous on its web site, “The membership wouldn’t exist in its present type with out his dedication, mentorship, and assist.” Morris usually attended conferences, wrote papers with scholar members, and suggested senior theses.
Along with his work on empirical Bayes, Morris made necessary contributions to chance principle. In a paper thought-about a breakthrough in statistics, he confirmed that there are precisely six households of chance distributions which can be members of the pure exponential household and have a quadratic variance operate. 5 have been well-known — Morris found the sixth to finish the set and derived many particular properties loved by these households. He made contributions to the hierarchical fashions broadly used at present within the social sciences, economics, and well being care coverage. Because the lead statistician for RAND’s well-known Well being Insurance coverage Experiment, Morris devised a brand new randomization scheme, the Finite Choice Mannequin, that was extra environment friendly than commonplace fashions and necessary to the examine’s success.
Morris held necessary management positions, most notably at Harvard the place he chaired the Division of Statistics from 1994–2000, years that weren’t with out controversy. The division was small however had internationally acknowledged students and a vibrant graduate program. The undergraduate focus was struggling, nevertheless, and had by no means had greater than 10 whole concentrators. There have been arguments inside and outdoors the division for dropping the statistics focus. Morris acknowledged the significance of statistics in shaping undergraduates’ understanding of the world past Harvard. Regardless of challenges, he believed in sustaining the focus throughout robust occasions. Because of Morris’s imaginative and prescient, the division now boasts internationally acclaimed packages and educates many Harvard college students, together with roughly 200 present concentrators.
Morris was a loving father and grandfather; he’s survived by his three youngsters and 5 grandchildren. He’s sorely missed by household, mates, and colleagues.
David Harrington, Chair