A Conversation with Monroe Sirken—His Early Career
Barry I. Graubard, Paul S. Levy, and Gordon B. Willis
This is an excerpt from the original article “A Conversation with Monroe Sirken,” published in Statistical Science in 2007. The interviews conducted by Barry Graubard, Paul Levy, and Gordon Willis took place in three sessions at the National Center for Health Statistics during the spring of 2006. Nearly the entire excerpt is from the initial interview conducted by Graubard on Monroe’s early career. Subsequent interviews on later stages of Monroe’s career were conducted by Levy and Willis.
Born January 11, 1921, in New York City, Monroe Sirken grew up in a suburb of Pasadena, California. He earned BA and MA degrees in sociology at the University of California at Los Angeles in 1946 and 1947 and a PhD in sociology with a minor in mathematics in 1950 at the University of Washington, where Z. W. Birnbaum was his mentor and thesis advisor. As a postdoctoral fellow of the Social Science Research Council, Monroe spent 1950–1951 at the statistics laboratory of the University of California at Berkeley and the office of the assistant director for research at the U.S. Bureau of the Census in Suitland, Maryland.
Monroe visited the Census Bureau at a time of great change in the use of sampling and survey methods and decided to remain. He began his government career there in 1951 as a mathematical statistician and moved to the National Office of Vital Statistics (NOVS) in 1953, where he was an actuarial mathematician and mathematical statistician. He has held a variety of research and administrative positions at the National Center for Health Statistics (NCHS), and he was the associate director of research and methodology and the director of the office of research and methodology until 1996, when he became a senior research scientist. He retired from that position in 2011.
Aside from administrative responsibilities, Monroe’s major professional interests have been conducting and fostering survey and statistical research responsive to the needs of federal statistics. His interest in the design of rare and sensitive population surveys led to the development of network sampling that improves precision by linking multiple selection units to the same observation units. His interest in fostering research on the cognitive aspects of survey methods led to the establishment of permanent questionnaire design research laboratories, first at NCHS and later at other federal statistical agencies here and abroad.
Monroe has been active in serving the statistical community. He has served on many committees of the American Statistical Association and the Washington Statistical Society (WSS). A charter member of the Federal Committee on Statistical Methodology (FCSM), he has chaired its research subcommittee that oversees a grants program in statistical and survey research that is funded by a consortium of federal statistical agencies and administered by the National Science Foundation. He is a Fellow of the American Statistical Association and the American Association for the Advancement of Science and is an elected member of the International Statistical Institute. He is a recipient of the Public Health Service Superior Service Award, the ASA WSS Roger Herriot Award for Innovation in Government Statistics, and the Federal Committee on Statistical Methodology distinguished service award.
Graubard: Then you went to the University of California, Los Angeles. Why that school?
Sirken: Well, I never considered going elsewhere. UCLA was virtually free for California high-school graduates with good grades. As I recall, UCLA tuition my first semester in September 1938 was $29 plus $4 for a student membership card that entitled me to admission to all UCLA sports events. Another financially related reason is, like most students attending UCLA at that time, I couldn’t afford to live on campus and UCLA was close enough to where we lived that I could commute. Foremost, I thought UCLA was a great university.
Graubard: I believe that your BA and MA were in the social sciences and I wonder how you became interested in statistics and mathematics.
Sirken: I earned my bachelor’s in sociology in 1946 and, the following year, my master’s in anthropology and sociology. How I became interested in statistics is a longer story. I began UCLA as a pre-med major with intentions of going to medical school, but during my sophomore year, I contracted tuberculosis. After recovering my health, about three years later, I returned to UCLA in 1943. In my financial situation, it was unrealistic to think of medical school. So, I changed my major to sociology, thinking that I might become a social worker. However, some faculty in the sociology department encouraged me to think about becoming a sociologist and advised me to take as much mathematics as possible. About 1945, I took my first course in mathematical statistics from Paul Hoel.
What have I been doing since retirement?
I officially retired from the National Center for Health Statistics at the end of January 2011. However, I continue to be listed on the NCHS roster as “expert without compensation,” which I believe gives me about the same status as being emeritus. So far, I have chosen to spend most of my time finishing projects that were interrupted by my retirement. Within the next few months, however, I’m planning to ease myself into a semi-retirement by taking courses offered by the community college and taking more trips offered by my senior residence community to theaters and nearby points of interest.
Advice to statisticians with recent PhDs
Whether you would prefer to work in the private, government, or academic sectors, work experience in all three sectors will broaden your knowledge and expand your professional networks and thereby help to advance your career. From personal experience, I can strongly recommend that you seriously consider government as the sector in which to start your career.
Advice for those currently working or interested in working in survey research
Because survey research is multidisciplinary—involving the behavioral, cognitive, computer, and statistical sciences—it will be to your advantage to have multidisciplinary training and experience in conducting interdisciplinary survey research. If your training is in the statistical and/or computer sciences, expand your knowledge in the behavioral and cognitive sciences; if your training is in the behavioral and/or cognitive sciences, expand your knowledge in the statistical and computer sciences. Positions at NCHS and other federal statistical agencies offer excellent opportunities to participate in interdisciplinary survey research projects.
Graubard: Why didn’t you stay at UCLA for your PhD?
Sirken: Well, for one thing, students were not encouraged to get the PhD at the same universities at which they were undergraduates. However, the more important reason is that I had become quite interested in quantitative sociology and UCLA did not offer that kind of graduate program at the time. I was awarded a fellowship in the sociology department at the University of Washington, which my UCLA advisors said was strongly quantitatively oriented.
University of Washington
Graubard: So what department did you actually end up in at the University of Washington?
Sirken: I started out as a teaching assistant in the sociology department. But one of the greatest strokes of good fortune in my professional life occurred after I arrived in Seattle.
In 1947, the math department at the University of Washington offered a two-year sequence of graduate courses in statistics, and the person in charge of the program was Professor Z. W. Birnbaum. I was unsure that I qualified, so I went over to (Bill) Birnbaum and described my background in math and statistics and my interests in quantitative sociology. To make a long story short, he accepted me and I took the courses he offered and did very well.
He was an excellent instructor who took an interest in his students. On the final day of the last class in the sequence, the students presented Bill with a gold-plated multicolor pen with the inscription “nature is not vicious” because that was the phrase he often used when the math got really complicated. When Bill died, his daughter told me that she found the inscribed pen on top of his desk.
In 1948, Birnbaum offered me a job as a research assistant in his newly established statistics laboratory, which I accepted. However, I had already accepted a position as a research assistant in the newly formed Washington Public Opinion Poll (WPOL), which was housed in the sociology department, where I was becoming acquainted with sampling and sample survey methods. So during 1948, I had distinct appointments in the sociology and mathematics departments, which I believe was quite unusual at the time and may have been illegal.
Graubard: I suppose that kind of interdisciplinary-type work would be encouraged now; back then, it must have been pretty unusual.
Sirken: Yes, I believe it was quite unusual in those days. I was just plain lucky to be in the right place at the right time. It happened because George Lundberg, chair of the sociology department, encouraged interdisciplinary research at the intersection of the social and mathematical sciences, and Z. W. Birnbaum was willing to take a chance on a social scientist.
I worked in Bill’s laboratory for two years and wrote my thesis under his direction. Bill taught me how to think in terms of statistical models. That was an invaluable gift.
Postdoctoral Fellowship and First Job
Graubard: What was your first position after getting the PhD?
Sirken: In the fall of 1950, I headed to the Statistics Laboratory at the University of California, Berkeley. I had a Social Science Research Council Postdoctoral Fellowship and was planning to spend most of the next 12 months at Berkeley. I took courses from Jerzy Neyman and Erich Lehman and consulted while there with Ed Barankin. I left Berkeley in June 1951.
Graubard: After working at Berkeley for eight or nine months, what did you do?
Sirken: I was pretty sure that I wanted to learn more about sampling and survey research, and what better place to get that kind of experience than at the U.S. Census Bureau? So I continued my fellowship there.
Graubard: Whom did you work with at the Census Bureau?
Sirken: Well, I was located in Morris Hansen’s office. The staff included really outstanding people such as Bill Hurwitz, Joe Daly, Max Bershad, and Margaret Gurney. These people were part of the central staff, and there were others also technically responsible to Morris who worked throughout the bureau, including Joe Steinberg, Joe Waksberg, and Harold Nisselson. However, I was mostly involved with a small group of survey methodologists, including Eli Marks and Leon Pritzker.
Graubard: What made you leave the Census Bureau and work at the National Office of Vital Statistics (NOVS)?
Sirken: It wasn’t my decision. There was a Reduction in Force (RIF) throughout the government soon after Dwight Eisenhower moved into the White House. I didn’t have tenure at the bureau and was RIF’d in 1953. A couple of months later, Morris Hansen got me a job at NOVS, which was quite a feat in view of the government-wide employment freeze. Well, Hansen and Halbert Dunn, the director of NOVS, were old friends and Hansen apparently convinced Dunn I was a competent actuarial mathematician. Later, when Dunn interviewed me, I noted my very limited knowledge (virtually none) in actuarial science, and luckily he thought I was being modest.
Graubard: What did you do after you finished the life table assignment, and who did you work with?
Sirken: Unlike the U.S. Census Bureau, where most national population data are collected periodically in censuses, at NOVS, national vital statistics (births, deaths, marriages, and divorces) are compiled as by-products of information reported on vital records. Because vital records serve primarily as legal documents, the information reported on vital records is necessarily limited and virtually changeless. I had an understanding with Dunn that, after completing the life tables, I would work on sample survey methods to improve vital statistics. Just after the life table project was completed, I had a chance to do just that when Bill Haenszel, a well-known epidemiologist at the National Cancer Institute (NCI), proposed a collaborative research project in which NOVS would design and test sample survey methodologies to collect retrospective residence and smoking histories for samples of deceased persons from their surviving relatives.
With funding support from NCI, a small statistical unit that included Mort Brown, Jim Pifer, and me was established in NOVS to conduct Haenszel’s pilot study. Soon after the successful completion of that pilot, NOVS established a long-range research sample survey program to expand the scope and improve the quality of vital statistics by conducting retrospective sample surveys linked to birth and death records.
Graubard: So you were using the idea that people like Hanson had promoted at the Census Bureau that by sampling you could expand the scope and improve the quality of vital statistics.
Sirken: The NOVS survey program was sustained by conducting work for other government agencies. For example, Haenszel expanded the lung cancer pilot study into a national mortality survey and arranged with the bureau to collect information on smoking and residence histories for the national population in the Current Population Survey (CPS). Thus, we were able to estimate national lung cancer death rates by smoking habits and residence histories. As I recall, these findings were cited in the first report of the Surgeon General on smoking and health.
Graubard: And actually that’s one of the first examples of a population-based case-control study, where the mortality follow-back survey provided exposure and other covariate information for the lung cancer cases and the CPS provided these variables for the control sample of the population at risk.
Sirken: Exactly. I have always felt that the linked mortality/population sample survey methodology deserves more attention than it has received from epidemiologists.
Graubard: Weren’t you also involved in designing other surveys for federal health agencies?
Sirken: Yes. NOVS developed the methodology of the follow-back surveys linked to birth records, and with funding from the Division of Radiological Health, U.S. Public Health Service, conducted the first national natality survey on the exposure of pregnant women to medical radiation. With funding from the U.S. Public Health Service, NOVS contracted with the Census Bureau for CPS supplement on the population’s utilization of the Salk vaccine. Using data from the CPS polio supplement, NOVS produced the first national statistics on the utilization and effectiveness of the Salk vaccine. Thereafter, the Public Health Service often used the CPS supplements to monitor the immunization status of the national population.
On another occasion, the U.S. Children’s Bureau asked NOVS to conduct a survey on the prevalence of cystic fibrosis, a debilitating and often lethal pediatric disease, and to do so within something like 100 days in order to comply with a congressional request. An unexpected estimation problem in that medical provider survey ultimately led to the development of a new kind of sampling called network sampling. These were very exciting days, when the findings of the NOVS sample surveys were used in real time to address important public health problems.
Graubard: When and why did you leave NOVS?
Sirken: In 1960, NOVS was merged with the National Health Survey (NHS) to form the National Center for Health Statistics (NCHS). Without any action on my part, I became a charter member of the NCHS.
Graubard: What kind of place was NCHS in those early days?
Sirken: A very exciting place. Forrest Linder was a good administrator and had the vision of developing a family of national data systems that intersected all important health-related activities of the population that would be capable of meeting the increasing needs for national health and vital statistics.
Download the complete version of “A Conversation with Monroe Sirken” to continue reading about Sirken’s innovative work on network sampling, integrated survey design, cognitive aspects of survey methodology, and the funding opportunity in survey research.