James Cochran

Peter (Tony) Lachenbruch

Tony Lachenbruch earned a BA in mathematics from the University of California at Los Angeles (UCLA) in 1958 and an MS in mathematics from Lehigh University in 1961. He then returned to UCLA and completed his PhD in biostatistics in 1965, after which he joined the faculty of The University of North Carolina, rising to the rank of professor in 1976. From 1976 through 1985, Tony was professor and head of the department of biostatistics in the department of preventive medicine and environmental health at the University of Iowa. He then returned to UCLA as a professor in the department of biostatistics in the school of public health, and he chaired that department during the 1991–1992 academic year.

At that point, Tony joined the U.S. Food and Drug Administration. While at the FDA, he served as chief of the biostatistics branch of the Division of Biostatistics and Epidemiology Branch in the Center for Biologics Research and Review through 1999, and then served as a senior biological research scientist and director of the Division of Biostatistics in the Center for Biologics Research and Review from 1999 through 2006. In 2006, Tony became a professor in the department of biobehavioral and population sciences at Oregon State University, the university from which Tony now holds the title of professor emeritus.

During Tony’s long career in biostatistics, his research interests have included statistical epidemiology and discriminant analysis, along with applications in rheumatology, cardiology, and psychiatry. Much of his efforts have been devoted to the development of statistical methods for drug research, selecting predictor variables when missing data are present, and regulation of pharmaceuticals. He also has worked on applications to rheumatology methods. Tony has authored or coauthored more than 200 articles and the book Discriminant Analysis.

Tony has received several important awards and honors. In 1971, he received the APHA (American Public Health Association) Mortimer Spiegelman Award; in 1991, he was named the Lowell J. Reed Lecturer by the APHA Statistics Section; in 1997, he received the FDA Award of Merit; and he received the FDA Center for Biologics Evaluation and Research Center Director’s Award for Distinguished Service in 2005. Tony was made a Fellow of the American Statistical Association in 1979, and, in 1983, he was elected to membership in the International Statistical Institute. In addition to serving as president of the ASA in 2008, Tony was president of ENAR in 1984 and WNAR in 2010.

Although Tony is now retired, he remains active professionally and continues to contribute to our discipline, serving on the editorial boards for Statistics in Medicine, Statistical Methods in Medical Research, and The Stata Journal.

In the second installment of the Amstat News series of interviews with ASA presidents and executive directors, we feature a discussion with 2008 ASA President Peter (Tony) Lachenbruch.

Q. Thank you for taking time to talk with me, Tony. During your childhood, were there any clues that you would become a statistician?

A. Not really, but I did enjoy mathematics. I didn’t take trigonometry until college, so my math development was a bit delayed. My entry into statistics was prompted by an ad in the American Math Monthly offering scholarships, so I contacted Frank Massey. I had a course in Math Stat as a senior at UCLA that was taught by Paul Hoel and Tom Ferguson from Hoel’s book Introduction to Mathematical Statistics. I remember that a senior math major fainted during the first exam and Tom and I carried her out of the room. I had already finished the exam, so I didn’t need to re-enter the classroom.

Q. You have worked on an incredible array of statistics projects. What is your most memorable project you have worked on as a statistician?

A.I have had a long-standing relationship with a group of rheumatologists—about 30 years—so that would be my most memorable project. The project began when I was at the University of Iowa, and it continued after I got to UCLA and then went to FDA. We were developing measures of improvement in juvenile dermatomyositis. There are six core measures that can be combined to give an index. Some of these are nonlinear and some are CART based. I learned about CART and propensity scoring through this project.

Q. What surprised you most about being president/executive director of the ASA?

A. The amount of support available from the ASA office. I have also been in leadership positions with other statistical groups that did not offer nearly as much support. I became president-elect when Ron Wasserstein was first appointed and learned a great deal from him—as we both learned simultaneously. Bill Smith was in his last year as executive director and passed on a lot of useful information. Bill had been in office for some time and knew the ropes quite well.

Q. What particularly humorous or unusual incident happened to you while you were president/executive director of the ASA?

A. Part way through my tenure, I received a letter from a colleague and friend asking about the status of a very large donation in her will. I referred it to Ron Wasserstein and the issue was then sorted out. This inspired me to make a donation to the ASA in my will.

Another less pleasant incident occurred when a person took offense at some ASA policy and wrote me in all CAPS telling me what rats the ASA and I were for not seeing things his way. After I responded a few times, I asked Ron what I should do and we came up with a solution: stop trying to explain the policy.

Q. What are your feelings about the future of the ASA? Do you have any particular concerns about the discipline of statistics or how the discipline is perceived by the general public?

A. The ASA is the greatest home and support for statisticians we have, and we must support it. We also need to increase the general public’s understanding of randomness. A large n means nothing if the sampling is biased. One of my great irritants are the ads that talk about a clinical trial that showed a drug was wonderful, not noting the number of trials that had been conducted. A single ‘significant’ trial out of 10 is not much to write home about.

Q. What misconceptions about statistics held by the general public would you like to see addressed?

A. More public information about issues in sampling and study analysis. Too many ads on TV talk about a study that showed something—but didn’t say how many studies didn’t show anything. Another issue I encountered while I was at the FDA was a study that changed the endpoint after the data were collected. The original study had used a nonparametric test and changed to a bootstrap test. There was a very large outlier in the control group that made significance in the bootstrap a sure thing, while the nonparametric test properly discounted it. The sponsors figured this out and were quite bitter when we didn’t allow them to change the endpoint.

Q. Is there anything that makes you particularly optimistic about the ASA’s future?

A. Yes, our younger members and their enthusiasm. I think these people will explain (much better than I have) these issues.