Samuel W. Greenhouse 1918-2000

by John M. Lachin and Joel Greenhouse

Samuel W. Greenhouse was born on January 13, 1918 in the Bronx, New York. Sam, as he was known to all, was one of the founding statisticians at the National Institutes of Health, who helped pioneer the use of statistical methods in epidemiological research, and was influential in the early development of the theory and practice of clinical trials. He was also a distinguished Professor of Statistics at the George Washington University.

Sam received his B.S. in Mathematics from the City College of New York in 1938 and thereafter moved to Washington, DC to begin his career in the Bureau of Census with Edward Deming (1940-42). He served in the Army during World War II and afterwards worked with the United Nations Relief and Rehabilitation Agency (1945-48). In 1948 he was recruited by Harold Dorn, along with Jerome Cornfield, Jacob Lieberman, Nathan Mantel and Marvin Schneiderman, to create the first biometry group at the National Institutes of Health (NIH) in the National Cancer Institute (NCI). The May 1997 issue of Statistical Science presents a description of the activities of this group including interviews with several of the early NIH statisticians, and Sam's own reminiscences and reflections on the development of statistics at the NIH. In 1954, Sam left the NCI to become Chief of the Theoretical Statistics and Mathematics Section in the National Institute of Mental Health. In 1966, he was appointed Chief of the Epidemiology and Biometry Branch of the National Institute of Child Health and Human Development, where he rose to the position of Associate Director for Epidemiology and Biometry (1970-74) and Acting Associate Director of the Office of Program Planning and Evaluation (1969-74). He was the first statistician to hold such a high administrative position at the NIH.

While working full time at the NIH, Sam taught part-time and pursued his own graduate degrees under the direction of Solomon Kullback in the Department of Statistics at George Washington University (GWU). When Sam retired from government service in 1974, he began a full time academic career at GWU where he served as Chair of the Department of Statistics from 1976-69 and again in 1985-86. In 1988, he retired from the University faculty and was named Professor Emeritus. From 1988 until his death he served as the Associate Director for Research Development of the GWU Biostatistics Center.

Sam articulated many times that the primary mission of the statisticians at the NIH was to collaborate and provide statistical support for the NIH scientists. Yet, it was always understood that these collaborations would lead to opportunities for statistical research in methodology and theory. It was not unusual to find Sam and the other early NIH statisticians co-authoring papers in subject matter journals and publishing corresponding theory and methods papers in statistics journals. This pattern was evident in his early papers on the evaluation of diagnostic tests. Although this work with Mantel (Biometrics, 1950) and Dunn (Public Health Reports) was rooted in the need to implement noninvasive methods for cancer screening, it also addressed methodological issues, such as deriving the estimated variance of sensitivity and specificity for the case when the diagnostic cut-point for a quantitative test was also estimated from the data. While at the NIMH, he helped design and analyze the first multi-disciplinary study of normal aging, and co-edited the resulting book Human Aging (1963). Recognizing the need for methods to analyze highly correlated psychological data from studies such as this one, led directly to new methodological work with Geisser (Psychometrika, 1959; Annals of Mathematical Statistics, 1958). They derived an estimate of the degree of departure from the assumption of compound symmetry in the test of within-subjects effects in ANOVA, and an adjustment to the degrees of freedom of the F-ratio when that assumption is violated. The Greenhouse-Geisser correction is now provided in virtually all computer packages for repeated measures analysis. Significantly, this work has been recognized as a Science and Social Science Citation Classic (July, 1982).

Sam was also influential in the early development of the theory and practice of clinical trials and shared an interest with Cornfield in methods for the sequential analysis of emerging data in clinical trials. While at the NICHD, his collaborations focused more on observational data, e.g., assessing the safety of oral contraceptive use, and his interests returned to the development of methods for epidemiologic studies. His papers with Seigel ( Journal of Chronic Disease, 1973; American Journal of Epidemiology, 1973), for example, showed that logistic regression could be applied to matched and unmatched case-control studies to obtain an adjusted estimate of the prospective odds ratio associated with a factor. At GWU in the late 1980's, Sam and Joe Gastwirth recognized similarities between a class of problems arising in legal settings and in epidemiologic studies. A collaboration began that was deeply grounded in the practical experiences of their respective fields of application (see, e.g., JASA, 1987; Statistics in Medicine, 1995).

Sam was passionate about statistics. He relished the opportunity to teach and engage colleagues and young statisticians in statistical discourse. Whether giving a seminar, making a site visit, or on sabbatical, Sam was always a popular and stimulating visitor. However, if one asked Sam about the truly important work he was doing, he would inevitably talk about his scientific collaborations. For it was through the practice of statistics, he believed, that statisticians made their biggest impact on science, and it was through scientific collaborations that the important statistical problems were identified. He felt that every biostatistician should spend time in the trenches, such as in a laboratory or a clinical trial data center, to obtain practical experience. He practiced what he preached. While at the Biostatistics Center he continued to collaborate extensively with investigators in the Division of Cardiology at the GWU School of Medicine; served as a coinvestigator of the Coordinating Center for the study of the Medical Treatment of Prostatic Symptoms; and served as a member of the data monitoring committee for the POSCH.

Sam was a much sought after panel member and mentor. His reviews and advice were always fair, insightful, and expertly crafted. He worked tirelessly as an advisor and reviewer for a number of government agencies, including, the U.S. Public Health Service's Accident Prevention Study Section (1958-62), the Federal Aviation Agency Council of Research Advisors, Office of Aviation Medicine (1959-65), and the Food and Drug Administration's Biometric and Epidemiologic Methodology Committee (1967-72) which he chaired from 1969-72. He served on the Institute of Medicine's committee investigating the health of Vietnam veterans, and co-authored a report for the White House on The Health Effects of Low-Frequency Electricity and Magnetic Fields. Sam was a member of numerous committees at the NIH, during and after his tenure there, including the Biostatistics Fellowship Review Panel (1961-69), the Statistics and Mathematics Study Section (1963-70), the NCI Epidemiology and Biometry Contract Review Committee (1967-73), the Computer Sciences and Biomathematics Study Section (1974-78), and the NHLBI Clinical Trials Review Committee (1983-86). In the 1980's and 90's, he frequently served as a member of the ad hoc NIH study section that reviewed statistical methods grant applications.

Sam was a much loved presence in the profession. He attended the annual meetings of ENAR, the Society of Clinical Trials, the JSM and the AAAS without fail, and the ISI as often as he could. He was amazingly current and had strong opinions on all matters. He was not shy about asking questions of speakers, especially when he didn't understand a point (or felt that they didn't), and it would not be unusual for the discussion to continue in the hall or even later via email until he felt the issues were resolved. This was true whether the topic was statistics, literature, music, politics, religion or sports. His intellectual curiosity was voracious. In his 1997 Statistical Science article on his reminiscences of the NIH he wrote about how the group (Cornfield, Halperin, Mantel, he, and others) would often argue quite publicly over lunch about matters statistical and otherwise. Although one of Sam's most endearing features was his personal warmth and smile, he could also be quite the provocateur. We fondly remember times in the late 1970's and early 80's, when Sam would visit Max Halperin or Nathan Mantel at the Biostatistics Center. Sam loved nothing more than a friendly spirited argument and Max and Nathan were always eager to comply. Sam was capable of arguing either side of an issue and often would, especially if it would get a rise out of Max or Nathan.

It is a tribute to his energy and enthusiasm for statistics that Sam received many honors for his intellectual and professional contributions. The American Statistical Association in 1993 recognized him with their prestigious Founders Award, and in 1997 videotaped a discussion with Sam as part of the ASA series of conversations with distinguished statisticians. Sam was a Fellow of the American Statistical Association, the Institute of Mathematical Statistics, the American Association for the Advancement of Science, and the Royal Statistical Society, and an elected member of the International Statistical Institute. He was also a Fellow of the American College of Epidemiology and of the Council of Epidemiology of the American Heart Association. In 1969 he received the Superior Service Honor Award from the NIH and in 1976 was named a Johns Hopkins University Centennial Scholar. In December, 1999, Sam was recognized by the Harvard Institute of Psychiatric Epidemiology and Genetics for his lifetime contributions to psychiatric epidemiology and biostatistics. Unfortunately, because of his illness, he was unable to travel to deliver the lecture that he had prepared in honor of this occasion.

Sam died of cancer on September 29, 2000 at the age of 82.