Skip to main content

On dissertation length (1982)

Walter L. Smith, recently deceased, was a longtime member and sometime Chair of the Statistics Department. He was also a raconteur with a rather decent portfolio. I remember a small story from probably ’82 or ’83. Wally recalled a time when he was sitting around with John E. Littlewood, and Abram Besicovitch, probably at the time that Besicovitch was the current Rouse-Ball Professor of Mathematics at Cambridge and Littlewood was the past holder of the chair. The question arose as to what was the shortest dissertation that they knew of. After the expected brief pause for thought, Besicovitch said he knew of one that was seven pages. Littlewood followed, saying he knew of one that was a single page. I found this just a teensy bit intimidating. (I eventually clocked in at around 140 pages.) As the years went by, and my own stockpile of unfinished business grew and grew, it occurred to me that they might not have been talking about just existing dissertations.

Mike Hoekstra (alumnus)

A Small Department with a BIG Heart (1989)

1989 turned out to be an usual year in many ways, but it still turned to be not the bad era to enter statistics as a profession. New ideas were bouncing and professors were excited about what they were doing. I wasn’t so sure about UNC when considering graduate school in statistics, until I found out about it through some yellow technical reports sent to my undergraduate department office. While discussing with one of my professors there, who strongly recommended UNC, despite the fact that UNC had just lost two of the applied statistics professors whom I might want to work with, it turned out that he was right. He was actually one of the visitors in the department in the early 1980s, and UNC department had quickly recovered with new recruits and more emphasis on interdisciplinary research, coupled with the founding of the National Institute of Statistical Sciences in the nearby RTP. The chairman of the department Stamatis Cambanis was great in so many ways, and at foremost he genuinely cared about everyone in the department, including everyone graduate student. One of his advices to graduate students right after qualifying exams was that you should seek a thesis topic that you were genuinely interested, as you’re going to work on it for many years. The Hotelling lectures were a big draw for graduate students, especially for the upper-level students, and the department colloquium were great, with many interesting talks (and treats after the seminars), in addition to almost weekly talks by visitors to the Center For Stochastic Processes. There were not so much interactions with the small OR department (except when later the MCMC fanfare came along), even if the graduate students were sharing the same building, but the social interactions with OR graduate students were very strong and much fun. In addition to classes taught by regular faculty members, which I think, were all very interesting, authoritative, and very beneficial later on, graduate students also benefited greatly with teachings from visitors. One of them was professor Sándor Csörgő, —-who was very energetic and I remember seeing him playing tennis in the apartment community court, and he might be the person who prevented me from becoming a Bayesian earlier (though the department as a whole then was not sympathetic to Bayesian at all), as I was struggling to understand the interpretation of confidence intervals in Bickel and Doksum’s Mathematical Statistics book, and he suggested me to read then just published Peter C. Fishburn article of The Axioms of Subjective Probability, which was totally unappealing to me. Another is the already world renowned Prof. Willem van Zwet, whose advanced class on asymptotic statistics was attended by many faculty and visitors, and our 1989 class were probably the only ones who took the class for credit— I remember we got a final problem of approximating the distribution of a sum of uniform random variables by the saddle point approximations, and we collaborated together to provide, I think, the best possible solution, by comparing the approximations with the true distribution, using a result from Walter L. Smith’s Stochastic Processes class, but each of us got a “p”, which may be the “average” grade that we were supposed to get as a class. It’s probably a loss of the statistics profession that Prof. van Zwet’s very organized and clearly handwritten written lecture notes never got published for wider appreciation. But the spirit of graduate students were great, working together, and always helping each other out, whether in classroom, or in daily life and social lives. Statistics was pure then, and we didn’t anticipate that we were at the beginning of an era of interdisciplinary and applied statistics. Statistics is now being married and embedded with not just OR, but also computer sciences, and many other subject matters as well. So 1989 may be an auspicious beginning after all, as statistics can now become enriched and be revitalized again!

John Lu (alumnus)

On Professor Shankar Bhamidi’s speed (2010)

Back in 2010, my freshman year, I took Shankar’s Measure Theory course. He delivered a quite enjoyable class, and right of the bat I was impressed by his speed of speaking and writing while giving a very good lecture!

This picture is from the second lecture of the class when Shankar he had already arrived at Karatheodory’s extension theorem… I vividly recall not being able to resist my let’s-doubt-everything Greek heritage and raised my hand to politely question: “Are we going a bit too fast?”. The class was fully submerged in the eerie trance you would typically expect from a Monday morning, first-year grad course in Measure Theory, but I am pretty sure I spotted a few supporting nods coming my way. After an ever so slight consideration, Shankar offered a friendly and disarming smile reassuring the class, that we were not and continued unphased (and likely amused).
I quickly realized that what he meant to say was: “we are not going fast YET”… My memory is a blur now, but I bet we ventured bravely into martingales and OST, with the last couple of lectures set aside for reviews… The funniest moment though, came a few weeks later, when caught in his own speed Shankar started saying “toin cosses” sending the entire class and himself in a laugh frenzy!!
I am sure we all have many great memories from our first semester…Super late homework nights at Hanes and graduate student house gatherings.

The second exhibit is from such a night when second year students Eric Lock and Susan Wei drink with us (me, James Wilson, and Jenni Shi) in solidarity 🙂

Stefanos Kechagias (alumnus)

On the retirement of Professor Norman L. Johnson (1982)

Professor Johnson used to go two or three times a day to collect the mail from all the faculty, staff, and students in the statistics department and deliver it to the department’s mailroom. This was because he daily received many letters and papers that were very useful for him to write and update his encyclopedic books and the Encyclopedia of Statistics. As many know, he was very knowledgeable and up-to-date on many topics in statistics and probability, and he accumulated a lot of material that he kept very well organized in his office in the Phillips building.
Due to his retirement from UNC in May 1982, he had to move offices and kindly asked several students to help him move his many books and papers from his old office to the new one in the Smith Building across Cameron Avenue, where statistics students had their offices and the OR department was at that time. After this change, and astonished by the enthusiasm and response of the students, he gave Ed Frees a twenty-dollar bill to be distributed among students, which we first declined, to his surprise, but later accepted and donated to the funding of the next department picnic.
He taught the last course of his academic career during the spring of 1982 (Statistical Theory II, STAT 135), and several of us who took that challenging course and its difficult final exam remember it as a real privilege and an honor.
Students celebrated Professor Johnson’s retirement from UNC by making different sports jerseys featuring a screen print of his face and signature.

Victor Pérez-Abreu (alumnus)

On the creation of the OR curriculum (1970)

As I recall, Prof. Nicholson from the stat dept spearheaded the establishment of the Curriculum in Operations Research at UNC in 1970 (or thereof). Prof. Smith may have been a member of the faculty council at that time. The story I was told is that he was present in the faculty council meeting to answer any questions from the council about starting this curriculum. One of the professors from the English department raised a moral objection to studying OR, since he had read that OR had its origin in making war time operations more efficient during WWII. The story goes that Prof. Walter Smith defended studying OR with his usual quick wit. He said “My esteemed colleague from the English department is quite right. However, it is also brought to my attention that all the war manuals on allies side were written in English. So how do we justify teaching and learning this language?” Needless to say the OR curriculum was approved by the faculty council! 

Vidyahar Kulkarni (current faculty)