Q: You served as president of both the Decision Sciences Institute and American Statistical Association. How did you maintain a level of activity sufficient to become president of both professional societies?

A: I have always felt my professional obligations as a university professor went beyond teaching, research, and service within the university and included service to organizations in my profession. Consequently, I was active in both the American Statistical Association and Decision Sciences Institute, the latter a relatively young organization at the time, with members primarily from the field of business administration.

Service as president in each organization required three years, one each as president-elect, president, and past president. Fortunately, my tours of duty were separated in time, with my presidency of the Decision Sciences Institute occurring from 1977–1980 and my ASA presidency some years later, from 1984–1986.

John Neter has been the C. Herman and Mary Virginia Terry Professor Emeritus of Management Sciences and Statistics at the Terry College, University of Georgia, since 1990. He taught there from 1975 until 1989. John previously taught at Syracuse University and the University of Minnesota. He earned his BS from the University of Buffalo, his MBA from the University of Pennsylvania, and his PhD from Columbia University.

John served as president of the Decision Sciences Institute from 1978–1979; he served as president of the American Statistical Association in 1985. He was also chair of the Section on Statistics of the American Association for the Advancement of Science from 1991–1992. John has been named a Fellow of three professional societies: the American Statistical Association in 1965, the American Association for the Advancement of Science in 1965, and Decision Sciences Institute in 1980. He received the Distinguished Service Award of the Decision Sciences Institute in 1981 and the Founders Award of the American Statistical Association in 1990.

Throughout his long career, John’s research interests have ranged from statistical sampling of accounting populations and response errors in sample surveys to statistical linear models. He is the co-author of Applied Linear Statistical Models, Applied Linear Regression Models, Applied Statistics, several monographs, and numerous journal articles.

Q: What led you and your coauthors to first write the books Applied Linear Statistical Models and Applied Linear Regression Models?

A: To answer your question about my motivation for preparing the Applied Linear Statistical Models text, I need to go back to early in the 1950s, when Bill Wasserman and I were teaching statistics to business students at Syracuse University. None of the available elementary texts motivated the statistical methods with interesting applications from the area of business administration. Bill and I therefore decided to write our own introductory statistics text for business students that would employ interesting applications from business administration and economics to motivate students to appreciate the usefulness of the statistical methods presented. The result was Fundamental Statistics for Business and Economics, which was published in 1956.

By about 1970, after several editions of the elementary text, Bill and I decided to prepare a second-level statistics text for business and economics students. In addition to emphasizing applications in business and economics, we also wanted to unify the discussion of analysis of variance, covariance analysis, and regression. In most of the existing second-level texts, these topics were treated as separate subjects. Hence, students would not understand the common structure underlying these statistical methods. Furthermore, the discussion of multiple regression and covariance analysis tended to focus largely on computational methods. It was apparent to us by 1970 that computers would easily enable statisticians to carry out linear model calculations and an extensive study of hand computation for linear models would not be required as a result. Hence, our presentation would focus on the nature of the linear models and their uses and on diagnostic methods, rather than on the hand computations required to invert a matrix.

As we began preparing materials for the second-level text in the early 1970s, it became clear that we could not adequately cover the range of topics we intended and also include ample illustrations of applications in business and economics without the book becoming much too long. As we realized a major reduction in the scope of the second-level book was required, we were quickly led to linear models as the primary focus of the book. The first edition of Applied Linear Statistical Models appeared in 1974. While preparing a second edition, the publisher decided to also present only the regression material in a separate book; this was published as Applied Linear Regression Models in 1983.

Q: In what volunteer roles had you served the ASA prior to being elected ASA president?

A: I served in many capacities prior to my election as ASA president. My very first service was in 1958 as president of the Twin Cities Chapter. And in 1962, when the National Statistical Meetings were held in Minneapolis, I chaired the local arrangements committee. In 1966, I headed the program committee for the National Statistical Meetings.

I also served on a number of committees of the American Statistical Association and two terms on the ASA Council. I was elected to the ASA Board of Directors for two terms, from 1974–1980, and served as editor of The Americans Statistician from 1976–1980.

Q: What were the main issues that the American Statistical Association faced during your term as president?

A: A key issue that faced us during my three-year term as ASA president-elect, president, and past president was acquiring our own office building. This was a major effort, involving site acquisition, building plans, fundraising, rental of unneeded space, and financing. The ASA had rented office space in downtown Washington as far back as I can remember, and the organization was quickly outgrowing this space. Also, we wanted the office to be located more conveniently for the many members from out of town who were attending the numerous meetings held each year at the ASA office. Ralph Bradley led the fundraising effort in the early 1980s and was very successful. Management had identified a suitable site and building in Alexandria and recommended proceeding with the purchase. I chaired a building acquisition committee and we examined the feasibility of proceeding. Unfortunately, we concluded that this was not the time to proceed. The real estate market in the Washington area was quite weak at the time, so we could not be assured of being able to rent the extra office space at a sufficient rental rate to break even; our financial condition did not permit us to risk operating losses. Fortunately, a few years later, we were able to proceed at a different site in Alexandria.

Another issue we faced at the time I was involved in the ASA leadership was the need to attract and retain applied statisticians. We considered, in particular, our publication policy and the annual meeting format to make the association more attractive to applied statisticians.

Still another issue was long-range planning. Historically, the American Statistical Association had not engaged in long-range planning. I asked Don Marquardt, president-elect, to undertake the ASA’s first long-range planning effort, which he carried out in most effective fashion. We were also very much concerned with encouraging the teaching of statistics in public schools and promoting greater and better use of statistics and statisticians by the sciences.

Q: Your textbooks have been important resources in applied linear models courses for many years. What are the most important trends in statistics education you have seen develop during your career? What are your thoughts about the current state of the discipline?

A: In my professional career, which ended 25 years ago, I was always interested in decision making based on good data. My interest was not only in the statistical methods used to aid decision making, but also in the relevant data needed for making good decisions—hence my interest in sample surveys and formally designed statistical experiments.

Thus, I have been delighted to see in recent years the increasing emphasis on data-driven decision making and the availability of large databases.

I am also pleased that the popular media are paying increased attention to the field of statistics and the contributions made by statisticians. By coincidence, the current issue of The Economist contains an interesting article about the impact of World War II on the field of statistics in Britain, and, in turn, the effect of statisticians on the war effort.

As in the days when I was active in the field of statistics, I note that statisticians today continue to be challenged to work with other disciplines so the full benefits of statistical methodology can be achieved.