Jim Cochran

Q: Sastry, thank you for taking time for this interview. When you began your undergraduate studies, was statistics your first choice for a major? If so, what motivated you to choose a statistics major at that point in your studies? Have you been happy with the choices you made?

A: “It was an acquired taste … my dad gave it to me!”

In our high-school days, immediately after the 10th grade, we needed to decide whether to go in the mathematical and physical sciences (MPC) direction, biological sciences direction (BiPC), or humanities and economic sciences direction (HEC). Professions that were encouraged heavily at that time included an engineer, a medical doctor, or a businessperson. Depending on your choice of future profession, you chose MPC, BiPC, or HEC as your specialization for 11th and 12th grades.

I chose MPC, thinking that I would be an engineer. During 12th grade, we took various entrance exams to enter engineering colleges such as the Indian Institute of Technology (IIT). In 1974, in addition to the engineering exams, I took an entrance exam for the Indian Statistical Institute (ISI) because my father was a mathematics teacher and one of his former students, Professor Hanurav, who was the dean of studies at ISI, talked to him about the benefits of going to ISI. These entrance exams are taken by 12th-graders from all over India. I recall gaining admission and getting mentally ready to study mechanical engineering at an engineering college. Then, I got a telegram from ISI saying, “You have been selected for an interview. Come prepared to stay if selected.” ISI was late in its selection process that year. About 100 students from all over India who did well on their entrance exam were interviewed, and after the in-person interviews, about 25–30 students were selected to attend ISI. I was fortunate to be one of the lucky ones selected to stay!

But I still had to decide whether I should go back and pursue mechanical engineering or stay there and study statistics. My father convinced me that I would have a better future in statistics than in engineering, and the rest is history. I am grateful to my father for opening the doors to the exciting profession of statistics.

A side story: After I became a full professor at NC State, my father did confess that he wasn’t sure whether he could financially support me to go to an engineering college on his salary in 1974.

I had five siblings, all going to college, and my oldest brother was already studying at an engineering college to be an electrical engineer. It wasn’t cheap to pay for room and board and other expenses for him. My other older brother and an older sister were in college at that time, as well. ISI, on the other hand, not only had no tuition charges, but it also paid us a stipend of Rs. 100 a month. And my hostel charges were only Rs. 10 a month. It was a great deal!

Q: So fortune smiled on you when you were choosing between studying mechanical engineering and statistics?

A: ISI is a great place to study statistics. Its motto is “Unity in Diversity.” It not only gave us a strong foundation in applied and theoretical statistics, but it also exposed us to various biological, computational, and physical sciences. I received truly a π-lingual training—depth in statistics and probability, breadth across other sciences for communication and collaboration, and solid computational skills. I graduated with BStat (Honors) and MStat and joined another statistical Mecca—Iowa State University (ISU) in 1979. Wayne Fuller and other outstanding ISU faculty provided me an excellent, well-rounded training in statistics. My luck didn’t stop there!

In 1982, I was fortunate to get an assistant professor job at NC State’s department of statistics, which was founded by Ms. Gertrude Cox. The department, which has distinguished alumni like Stu Hunter among many others, was the birthplace of SAS and journals such as Biometrics and the Journal of Statistics Education. The statistics department has also appreciated interdisciplinary research since its origins. I had a wonderful time growing up in my professional career at NC State for 30 years—moving up the professorial and administrative ranks as the director of graduate programs and the head of the department of statistics. I loved my job, my students, my alumni, and my colleagues at NC State. As Ron Wasserstein would say, “If I get a penny for every time Sastry says ‘NC State is one of the oldest, largest, and the best statistics departments in the country,’ I will be a rich man.”

So, I have been very lucky with ‘my’ choices in statistics, and as my wife, Sobha, and daughter, Asha, would say, “You better stick to budgets. We will call someone to fix that door knob.” Clearly, I am no mechanical engineer!

Q: Besides your theoretical work in time series analysis, are there any particular projects you worked on that were most interesting or challenging?

A: The majority of my theoretical work focused on testing for nonstationarity of time series and on inference in nonlinear random coefficient models. Some projects I really enjoyed working on with my colleagues and others do not appear in journals. For example, I thoroughly enjoyed my sabbatical at a semiconductor consortium, SEMATECH, in Austin, Texas. I had an opportunity to work with many engineers on well-designed experiments related to the quality, reliability, and capability of thin wafers. I enjoyed developing and delivering one-day and three-day workshops on quality control, capability analysis, and experimental design throughout the country and in Mexico.

I also enjoyed working on classifying the dives of seals in Alaska, whether the dives are V-shaped, U-shaped, or bathtub-shaped. Each shape indicates their activity, and more importantly, indicates the location of fish where seals are prone to benthic eating. Similarly, I enjoyed my work with colleagues related to home-range estimation of animals and work on multi-frame estimation of population counts from combining various administrative records.

I have the highest regard for my colleagues John Rawlings and David Dickey, with whom I had the pleasure of collaborating on a book, Applied Regression Analysis: A Research Tool (2001). I am certainly grateful and lucky to have had such colleagues.

Q: You served as director of the National Science Foundation’s Division of Mathematical Sciences (DMS) from 2010–2013, a position for which you have a unique background. Had any other statistician(s) served in that role prior to you? What challenges did you face in this position because of your unique background?

A: No, I don’t think any other statistician has served in the role of the DMS director.

NSF is a wonderful place to work, and I highly recommend it. It is a myth to think that government workers are lazy and do not appreciate the proposals you submit. On the contrary, the program officers at DMS are very hard working and labor hard to stretch dollars to support outstanding research. It is an honor to work with folks at NSF. The main challenge is the shrinking budget and increased demand for research dollars. There is certainly a significant amount of unfunded excellence and grants simply because there is not enough money to support all excellent research proposals.

As director of DMS, my challenge as a statistician was not much different from what a topologist or probabilist or applied mathematician would face in the same role. We are responsible for all of our programs, regardless of where our heart may be. Of course, it was challenging to see the number of statistics proposals growing at a faster rate than funding for the statistics program or DMS in general. Overall, this continues to be an issue, especially given that research is more and more computationally and data-enabled. Statistics program officers are also challenged and stretched to connect with every division at NSF in the same way that statisticians collaborate with all sciences, engineering, and education. Statistics research will continue to grow, and hopefully this will be reflected in the DMS budget.

Based on the growth of statistical sciences and departments of statistics during the past 50 years, I did initiate a proposal to change the name of DMS to the Division of Mathematical and Statistical Sciences, which was not successful. However, the proposal did generate much discussion among statistics communities and within NSF about the role of statistics and, more broadly, about data science. I enjoyed working with new programs at NSF such as the Computational and Data-Enabled Science in Mathematical and Statistical Sciences (CDS&E-MSS), Science Across Virtual Institutes (SAVI), and BIGDATA. I found it satisfying to recruit and work with diverse groups of program officers and staff.

Q: You spent nearly 30 years as a faculty member with the statistics department at North Carolina State University (NCSU), during which time you were inducted into the NCSU Academy of Outstanding Teachers. You then moved into administrative roles, first with the NSF and then as dean of the Oregon State University College of Science. What motivated you to move from teaching into administration? What do you miss about teaching?

A: Teaching is truly in my blood. ‘Pantulu,’ in my language, means teacher, and my last name, ‘Pantula,’ is derived from this word.

“Teaching is truly in my blood. ‘Pantulu,’ in my language, means teacher, and my last name, ‘Pantula,’ is derived from this word.”

My father was a math teacher, my sister is a teacher, my maternal uncles and aunts are teachers, and my grandfathers were teachers and writers. My father used to wake me up at 4 a.m. to teach calculus to several of my friends and me in our living room. He loved to teach, and he literally taught me to love teaching. I used to tutor high-school kids when I was at ISI, and I volunteered to teach at ISU one summer. Seeing students understand new topics really excites me. So, I am certainly passionate about teaching and teaching as a profession. Even with reduced teaching loads in my administrative roles at NC State, I still was able to enjoy teaching students who were preparing for qualifiers or master’s comprehensive exams. Some used to come to our house on weekends, and others used to meet me after hours in SAS Hall. My favorite subjects to teach are linear models, time series, and undergraduate probability and statistics for engineers. I have not had an opportunity to teach at Oregon State yet, but I did have an opportunity to read and discuss a book, Numbersense, with students from the university honors college.

First, the love of working with students and wanting to make an impact on increasing the quality, quantity, and diversity of students excited me to take on the role as director of graduate programs in 1994. I worked closely with the department head and a close friend of mine, Tom Gerig, and eventually added on the role of associate head. This piqued my interest in becoming a department head in 2002 in order to make even more of an impact by recruiting outstanding diverse faculty and making NC State’s statistics department more recognized globally for its excellence and diversity. I was very fortunate to have harmonious faculty and a supportive dean to help our faculty and students achieve many common goals.

Q: It is plain to see that teaching really is in your blood. How have your professional activities allowed you to continue contributing to education?

A: I was incredibly lucky when I was nominated for the 2010 ASA president position, and felt very honored when our members elected me to this office. It gave me a broad perspective of our incredible profession and the diversity of sectors the ASA represents. I appreciated the value and need to grow our membership, to show the impact of our profession, to be visible to policymakers, and to help educate and train future problem-solvers. GIVE to ASA was my theme—Growth, Impact, Visibility, and Education. I really enjoyed getting to know our members and seeing firsthand how dedicated they are to promoting the practice and profession of statistics.

Being the division director of DMS at NSF gave me broader exposure to other sciences, to engineering, and to education. This position also provided me with an invaluable opportunity to work with others to lead change at a national level. But as my tenure at NSF was ending, I considered seriously going back to teaching and research at NC State or George Washington University. I was evaluating myself, specifically where I could contribute the most and take advantage of the fortunate and wonderful experiences I had. Oregon State University (OSU) presented an extraordinary opportunity as dean of the college of science to move the dial AHEAD—Advancing Harmony, Excellence, And Diversity—three things that are close to my heart. People here are very humble, kind-hearted, and excellent in everything they do. Here at Oregon State, I am fortunate to have a fantastic administration and colleagues, a great network of alumni and supporters, and staff who like to see everyone succeed. We build leaders in science at OSU.

Like teaching, I realize that service to the profession, playing a small role in others’ successes, and helping people reach their full potential is extremely gratifying. Ultimately, it is this selfish gratification in service that moved me into administration. I am fortunate, and I love it.

What I most miss about teaching is the interaction with students and the adrenaline rush I get being in front of a class.

Q: What challenges do you feel the ASA and our discipline will confront in the near future?

A: The ASA is a like a big tent, a very inclusive place. Members from various sectors and different backgrounds respect each other. The ASA proactively looks out for the health and best practices of the profession. Also, very importantly, it plays a role in policymaking for the country and in the development of the future workforce.

For all the things it does or plans on doing, the ASA will need to have a solid financial base. In addition to membership, publications, and meetings, it should enhance its fundraising activities. Our leadership, industry partners, and philanthropic organizations can continue to help the ASA achieve its goals. GIVE to ASA!

The ASA has successfully moved the profession away from the “we get no respect” attitude to a prominent place at the table, where we can add our voice to a national dialogue and inform important decisions. A good example is the work with AAAS and Science magazine. Statisticians are at the table! We need to deliver on our promise to build trust in science. The ASA has incredible staff, but they may be spread too thin to keep an eye out for so many essential areas such as statistical agencies, teacher training, academic leadership, and the health of statistics departments in universities, as well as isolated statisticians.

As a discipline, we need to help diversify our workforce and pay attention to under-represented minorities. It requires efforts at all levels to be prepared for the changing demographics in the United States. Also, we need to be partners with other sciences to play a key role in Big Data and analytics. As data become ubiquitous, statistics becomes more and more relevant. For instance, precision agriculture, precision engineering, and precision medicine all need statistics. Whether we call it data science or statistical science, at the heart of it is fundamental statistics and good ethics. We cannot afford to have a blind spot in a computationally and data-enabled world. This means training future leaders in statistical science!

I remain very optimistic for our profession and for the ASA! I am incredibly honored and lucky to have the opportunity to serve.

Please return to this column next month, when we will feature an interview with 2004 ASA president, Bradley Efron.