Emily Palmer, Harvard Department of Statistics Administrative Coordinator writes:

On May 5, 2023, Professor Joe Blitzstein from the Harvard University Department of Statistics launched the centennial celebration of emeritus faculty Herman Chernoff, and his research and teaching legacy. The day featured a full program, including a tribute video from former colleagues and students; an interview with professor Xiao-Li Meng; and research presentations from professor emeritus Joseph (Jay) Kadane of Carnegie Mellon University, professor Tian Zheng of Columbia University, and professor Joseph Gastwirth of The George Washington University. The event recognized Chernoff ’s contributions as a statistician, educator, and mentor.

Herman Chernoff, who turned 100 on July 1, 2023, started his career in 1943 with a BS in mathematics and a minor in physics from City College in New York. For a year and a half, he worked as a physicist with the US Navy, building and fixing electronics. According to John Bather in “A Conversation with Herman Chernoff,” published in Statistical Science, Chernoff’s use of statistical ideas in the Navy convinced him to return to school to pursue a master’s and PhD in applied math at Brown University, where he was supervised by Abraham Wald. Chernoff held faculty positions at the University of Illinois (1949–52), Stanford University (1952–74), MIT (1974–85), and Harvard University (1985–1997).

Chernoff ’s contributions to the field include work on large sample theory, experimental design, sequential analysis, presenting statistical data in visual form, and statistical decision-making. He is also known for his enthusiasm for mentoring. He proposed creating the annual New England Statistics Symposium to support young researchers and, in honor of his contribution to the symposium and profession in general, the New England Statistical Society established the Chernoff Excellence in Statistics Award in 2019.

In recognition of his work, Chernoff has received honors from the National Academy of Sciences and American Academy of Arts and Sciences and been elected fellow of the American Statistical Association and Institute of Mathematical Statistics.

Herman Chernoff (L) interviewed by Xiao-Li Meng. Photo by Sally Thurston.

During a conversation with Xiao-Li Meng, Chernoff shared why he changed focus from mathematics and physics to statistics. He described a pivotal moment in his graduate student career when he read a paper by Wald about generalizing the testing of hypotheses and estimation. “Wald’s paper struck me because […] it confronted the fundamental idea that the test of a hypothesis or an estimation of a parameter leads to a conclusion and that conclusion should have an economic, real-world consequence.” He concluded, “And that’s what converted me to being a statistician!”

Chernoff added that reading papers by Jerzy Neyman and Karl Pearson exposed him to the idea that it was important to consider alternatives to a hypothesis when evaluating it. Those experiences highlighted what drew Chernoff to statistical thinking: his interest in connecting theory to applications and solving and quantifying problems of uncertainty. Although he was trained as a mathematician and physicist, Chernoff relished the opportunity to tackle a new field. In his career as a statistician, Chernoff embraced working in applied and theoretical areas, a rare feat today because of how specialized statistics has become. Blitzstein emphasized Chernoff ’s multi-faceted work in statistics, pointing out his applied and theoretical, Bayesian and frequentist, and parametric and nonparametric approaches.

The research talks given by Chernoff ’s colleagues showcased the span of his research interests and influence. While Kadane’s talk focused on using probability theory to analyze handwriting in a court case, Zheng’s talk covered methods for detecting influential variables in high-dimensional data, specifically genetic data. Additionally, Gastwirth spoke about his collaboration with Chernoff on the use of L-statistics to measure economic inequality.

In his interview with Meng, Chernoff explained why the manifold applications of statistics have motivated his career. He reflected, “People regard me as a theoretical statistician, but I’ve decided in recent years that I’m really an applied statistician. My theoretical insights have relied upon my work in thinking about applied problems.”

An example of Chernoff taking inspiration from applied problems was when he created Faces, a data visualization tool he developed to help researchers analyze multivariate data (by presenting data as faces), while at Stanford.

The centennial celebration illustrated Chernoff’s love of learning. The curiosity that drove him to study statistics also motivated him to study R. Blitzstein said, “I was pretty impressed that he was still coding. Usually, brilliant mathematicians get their PhD students to do all the coding, but Herman wanted to test out his methods by carrying out the simulations himself.” He added, “One day, Herman came into my office and asked for a book on C because R was ‘too slow’ for him, and I was even more impressed with that!”

Having taught statistics for many years, Chernoff still encouraged and mentored undergraduates once he became an emeritus faculty. Blitzstein reminisced about when their offices faced each other and he would routinely return to his office to find Chernoff in conversation with a student. “My students from Stat 110 would receive homework help from this friendly man in his 80s without even realizing he had pioneered some of the methods they were using 60 years before,” marveled Blitzstein. Chernoff demonstrated his interest in supporting students’ intellectual growth by publishing books such as Elementary Decision Theory in 1959 and Algebra I for Students Comfortable with Arithmetic in 2001. Holding his autographed version of the former, Blitzstein commended the book for being accessible to high-school students and addressing some of the most pressing questions in statistics.

Some of the most memorable moments of the centennial celebration were when colleagues and former students shared their stories about Chernoff ’s tips and coaching. For example, when Gastwirth was a junior colleague at Stanford, he was thrilled to be asked to review an article and willing to offer a tight turnaround with his comments. He shared advice from Chernoff that he’s adhered to ever since. “When I told Herman about finishing the review within a week, he said, ‘No, no—you have to think about the worst possible thing that can happen to delay you (e.g., you get sick), and then you double that amount of time!’”

Harvard statistics professors Meng and Jun Liu shared anecdotes about Chernoff ’s mentorship. Shortly after Liu arrived as a junior faculty at Harvard, he was drafting the first paper in which he was a first author and Chernoff offered to read it. Liu recalled, “I felt a little intimidated by such a renowned statistician offering to read my paper, but Herman read it and provided such useful, line-by-line comments that it was ultimately accepted into the Annals of Statistics.”

During his interview with Chernoff, Meng highlighted an episode that occurred early in his career as a PhD student. In preparation for his qualifying exam, Meng printed a copy of his paper to read to his committee. Chernoff paused the perusal with a question Meng glossed over. Chernoff interjected, “Xiao-Li, you are not answering my question.” Meng learned it wasn’t sufficient to pursue important research questions; he had to learn how to communicate effectively in order to succeed as a researcher.

During a tribute video and toasts, friends, family, former colleagues, students, and post-docs acknowledged Chernoff’s impact on them. They reminisced about grad student lunches at a pizza parlor and cozy gatherings at the Chernoff home, and they shared some of Chernoff ’s favorites: politics, travel, dogs, and Swiss orange chip ice-cream. Many also paid tribute to the late Judith Chernoff by reflecting on her warmth and humor and lauding the couple’s marriage of more than 75 years.

Throughout the celebration, there were many moments that brought Chernoff ’s sense of humor to the foreground. An example was when he narrated a story about his analysis of the Massachusetts lottery. When he arrived in Boston, the lottery had only been in place for about 500 days. After a statistician showed the lottery was likely to have had some repetitions but didn’t, Chernoff realized the lottery was probably fixed. He elaborated, “When I realized that the lottery was fixed, I thought that I should announce it, but then maybe the gangsters would not like that and would assassinate me. On the other hand, when I told the area chair at MIT, Harvey Greenspan, he suggested that if it was fixed, I should take advantage of it!”

Chernoff ’s words are emblematic of his approach to life and relationships with others. Despite his serious career pursuits, he appreciates humor and enjoys interacting with friends, family, and colleagues.

See more photos from the event at https://statistics.fas.harvard.edu/news/harvard-department-statistics-celebrates-professor-herman-chernoff%E2%80%99s-100th

—

This report first appeared in the Nov. 2023 Amstat News.