Stephen Fienberg

The National Institute of Statistical Sciences (NISS) Board of Trustees established the Jerome Sacks Award for Cross-Disciplinary Research in 2000 to honor Sacks’ service as the founding director of NISS. The annual prize of $1,000, presented at the NISS JSM Reception, recognizes sustained, high-quality cross-disciplinary research involving the statistical sciences.

Stephen Fienberg, Maurice Falk University Professor of Statistics and Social Science and co-director of the Living Analytics Research Center at Carnegie Mellon University, was honored with the 2015 Jerome Sacks Award for Cross-Disciplinary Research for “a remarkable career devoted to the development and application of statistical methodology to solve problems for the benefit of society, including aspects of human rights, privacy and confidentiality, forensics, survey and census-taking, and morte; and for exceptional leadership in a variety of professional and governmental organizations, including in the founding of NISS.”

Jamie Nunnelly, NISS’s communication director, conducted the following interview.

What got you interested in the field of statistics?

My interests in statistics date to undergraduate courses at the University of Toronto. In my third year, Don Fraser introduced me to the mathematics of statistics and inference issues. Much of what he did was explicitly geometrical, and that had great appeal to me. He also exploited the magical inversion associated with Fisher’s fiducial argument, wherein the distribution of the data given the parameter induces a distribution on the parameter given the data. In my fourth year, Dan DeLury taught a course on the design of experiments, and that led me to read Fisher’s book on the topic. At the same time, I took a research methodology course for psychologists taught by the eminent cognitive researcher Endel Tulving. There I could see how to put much of what I was learning from Fraser and DeLury to work. At that point, I was hooked and began to apply for graduate study in statistics.

Who were some of your influencers?

The department of statistics at Harvard, where I did my graduate work, was quite small, and every faculty member influenced me in some form. Paul Holland was a recent PhD from Stanford, and we worked on things motivated in part by Stein’s inadmissibility results, but from a Bayesian point of view. Art Dempster stimulated me to think about foundational inference issues, and Howard Raiffa and John Pratt, who were essentially in the business school, helped to reinforce my Bayesian tendencies. George Tiao, who visited Harvard, also played an important role in my Bayesian education and was an early coauthor. Bill Cochran and Fred Mosteller (who was my thesis adviser and later mentor) were my role models when it came to making statistics work in real applications, including matters of serious public policy.

Later, when I joined the faculty at The University of Chicago, Bill Kruskal and Paul Meier continued to draw me into diverse areas where statistics could make an impact. Even though Bill and I differed on matters of inference—I was already very much a subjective Bayesian and he was a committed frequentist—he represented for me the statistician as a public intellectual and he reinforced in me the importance of careful scholarship, not just the development of new methodology.

What was one of the first projects you did with the National Institute of Statistical Sciences (NISS)?

Actually, my earliest interactions with NISS came via service on the board, which I joined almost at the outset and on which I served for more than a decade. I pushed quite hard for a focus on developing the kinds of research projects that statisticians would typically not happen upon on the campuses of their own universities. This was before the era of Big Data, but clearly in the spirit of such. And, of course, the watchword was “interdisciplinarity.”

What was your favorite project you did with NISS?

In the 1990s, I had been drawn into working on the statistical problems associated with confidentiality and privacy protection. When Alan Karr and I attended a National Science Foundation (NSF) workshop launching NSF’s digital government initiative, we realized it would be an ideal vehicle to pursue confidentiality and privacy issues of interest to government statistical agencies and develop new methodology drawing together experts from a number of university campuses. The successful proposal we wrote went in from NISS, and it solidified my links to NISS as well as my friendship with Alan. The students and postdocs trained under our grant have gone on to make major contributions to this area of research.

What are you working on now?

I continue to work on a multiplicity of statistical problems, all of which involve Big Data and interdisciplinarity: confidentiality and privacy, including record linkage; methodology for census taking; forensic science and the law (including my involvement with the new Center for Statistics and Applications in Forensic Evidence—CSAFE); and network modeling.

What advice would you give someone who is thinking about entering the field of statistics?

I’m reminded a bit about the scene in the 1960s movie The Graduate, where Dustin Hoffman in the title role is offered one word of advice: “Plastics!” My one word of advice to someone just entering the world of statistics is “applications.” While statistics has an intellectual core built around probability and inference, I have always drawn my inspiration from real-world problems arising in other disciplines, and that is, of course, where the data we analyze arise. So I tell my students to take applications seriously and to use them to motivate the methodological and theoretical work they choose to do.

Anything else you would like to add?

Statistics is an amazing and challenging field. The opportunities today are limited only by our collective imagination. In some ways, it was serendipity that led me into statistics more than 50 years ago. The ever-expanding interest in statistics by our undergraduates and the demand for statisticians in government, academia, and industry reinforce the importance of thinking and working across boundaries. That is what the NISS Sacks Award is really about and why I was so honored to receive it.