Home » Member News, Obituaries, People News

Obituaries for April 2022

1 April 2022 No Comment

William Michael (Mike) O’Fallon

William Michael (Mike) O’Fallon

ASA Fellow, Founder, and president in 2000, William Michael O’Fallon died on February 19, 2022, in Rochester, Minnesota.

Born March 7, 1934, O’Fallon attended St. John’s University (SJU) in Collegeville, Minnesota, graduating in 1956. Since he intended to be a high-school teacher, he then studied at Vanderbilt University, where he earned his master’s degree in mathematics in 1957. However, he was recruited to teach mathematics at SJU in the fall of 1957.

In 1960, O’Fallon was accepted into the mathematical statistics doctoral program at The University of North Carolina at Chapel Hill, where he met his wife, Judith Rich. He completed his PhD in statistics in 1967.

O’Fallon worked as associate professor at Duke University for several years before taking a sabbatical to study epidemiology. In 1974, the Mayo Clinic’s department of statistics and epidemiology requested he design a breast cancer epidemiology study and, later that year, the department recruited him to chair the biostatistics section. O’Fallon held that position for 20 years.

During his time at the Mayo Clinic, O’Fallon worked as a professor in the Mayo Medical School, coauthored more than 450 peer-reviewed publications, and served as principal investigator on nearly 50 major grants. He was appointed to chair Mayo’s department of health sciences research in 1994, a position he held until 2000.

O’Fallon was elected fellow of the ASA in 1986 and he served as ASA president in 2000. In 2008, when asked to reminisce about his term as ASA president, O’Fallon responded, “It was an honor I never expected and was not at all sure I wanted. Indeed, when I told one of my colleagues at Mayo that I had accepted the nomination, she responded, ‘You must be certifiable.’”

O’Fallon went on to say, “The theme for my presidential year was that we as a profession, needed to embrace our diversity by recognizing and valuing the many talents statisticians of all levels bring to the table.” He also expressed his hope that the ASA would make progress toward embracing diversity and that the climate of togetherness would continue into the 21st century.

In 2002, the ASA presented O’Fallon with the prestigious Founder’s Award for “extraordinarily effective leadership.”

O’Fallon was committed to helping others. In retirement, he served as the interim director of the Channel One Food Bank. He was active in The Rotary Club of Rochester, serving as club president in 2005 and assistant district governor from 2009–2011. He was also active in his church, serving on committees, participating in men’s group activities, mentoring new members, and lecturing.

A memorial service to celebrate O’Fallon’s life will be held in Rochester this summer, when family members and friends can attend. In lieu of flowers, donations in O’Fallon’s name can be made to the Channel One Food Bank or your favorite charitable organization. Read O’Fallon’s full obituary online.

Michael Woodroofe

Michael Woodroofe, Leonard J. Savage Professor Emeritus of Statistics at the University of Michigan, passed away February 22, 2022, at the age of 81.

Michael made many fundamental contributions at the interface of probability and statistics, including sequential analysis and nonlinear renewal theory, shape-restricted inference, and central limit theorems under dependence. He was one of the founders of the Michigan Department of Statistics in 1969 and the last solo editor of the Annals of Statistics

Michael was an exceptional statistician and a great mentor who supervised more than 40 PhD students and a wonderful human being who touched many lives directly. 

Thomas B. Jabine

Submitted by Richard Valliant

Thomas B. Jabine

Thomas B. Jabine, an ASA Fellow since 1965 and an elected member of the International Statistical Institute (ISI), died on January 30, 2022, at the age of 97.

Tom was notable for a long and distinguished career in the US federal government, his service as a statistical expert to international government agencies, and his service to the ASA.

He earned a BS in math and an MS in economics and statistics at the Massachusetts Institute of Technology in 1949 and married Marian Brandon Smith, his wife of 56 years, in 1950. He was chief of the Statistical Research Division for the US Census Bureau, chief mathematical statistician for the Office of Research and Statistics in the Social Security Administration (SSA), and statistical policy expert for the Energy Information Administration (EIA). He was also a consultant to the United Nations; US Agency for International Development; Organization of American States; and the governments of Uruguay, Nepal, Iran, Turkey, Afghanistan, Sri Lanka, Bangladesh, and Papua New Guinea. Additionally, Jabine was the editor and author of several books on statistics and survey methods.

During his 20-year career at the Census Bureau, Jabine rose from mathematical statistician to chief of the Statistical Research Division. He was among the cadre of statisticians working for Morris Hansen and William Hurwitz who developed many of the methods in probability sampling that became industry standards. He branched out from sample design to questionnaire design after it became clear to him that minimization of error in surveys required work on both sampling and nonsampling error. During the 1950s and 1960s, he designed some of the first split-panel tests in government surveys of alternative questionnaire features and data-collection methods.

Soon after Jabine joined SSA’s Office of Research and Statistics, passage of the Privacy Act of 1974 imposed several new requirements on federal agencies collecting and disseminating data about individuals. He was charged with developing procedures for meeting these and other legal requirements at SSA and later at the EIA. In the 1980s, he was a member and chair of the ASA’s Committee on Privacy and Confidentiality. In addition to being an ASA Fellow and elected member of ISI, Jabine’s honors include serving as president of the Washington Statistical Society in 1978–1979 and receiving the Roger Herriot award in 1999.

At the Committee on National Statistics (CNSTAT) of the National Academy of Sciences, Jabine coedited or consulted on 13 of the committee’s reports. Among them were the 1984 report, Cognitive Aspects of Survey Methodology: Building a Bridge Between Disciplines, and the 1993 report, Private Lives and Public Policies: Confidentiality and Accessibility of Government Statistics. Research on cognitive methods in surveys and protection of the privacy of respondents to surveys and censuses has burgeoned since then.

Jabine felt some of his most rewarding work was providing technical assistance in census and survey methods to statistical offices in developing countries and later working on the role of statistics in human rights. In the late 1950s, he helped Brazilian census planners introduce sampling methods in their 1960 population census. From 1963–1965, he assisted the Thai statistical office in designing and initiating household sample surveys. In the 1980s, he served as a chair of the ASA’s Committee on Scientific Freedom and Human Rights. Initially, the committee’s efforts were directed toward helping statisticians in other countries whose rights were being violated.

Later, they extended the scope of that work and the committee’s charge to cover the use of statistics to monitor the status of human rights.

An interview with Jabine was done by Michael Larsen (2008) for CHANCE, for which Jabine was the puzzle master for 10 years. He was a co-founder of the Senior Statistician’s Society in Washington, DC.

Survivors include his four children, Thomas P. Jabine of Washington, DC; William T. Jabine of Buffalo, NY; Anne (Angie) B. Jabine of Portland, OR; and Leslie N. Jabine of Chicago, IL; seven grandchildren; and two great-grandchildren.

Charles R. Mann

Lee-Ann C. Hayek and Richard Sampson

Charles Mann—president of Charles R Mann Associates in Washington, DC; ASA Fellow; and longtime ASA member—passed away on February 14, 2022, after a long illness.

He had been a member of the ASA since the late 1970s and, between 1980 and 1986, served on the Ad Hoc Committee for Ethics, which initiated and structured the original code of ethics. He was a member of this committee when members requested the ASA Board to establish a permanent ethics committee with the purpose of developing a permanent ethics code for the association. Charles was a member of this committee from its beginning in 1986 until the acceptance of the code of ethics in the early 1990s. He—along with T. Ireland, the first chair of the ad hoc committee, and L. C. Hayek, cochair of the ethics committee—wrote the code, and the latter presented it to the ASA Board, obtaining acceptance. Charles considered this among his highest achievements for the ASA.

Charles was also a member of the Washington Statistical Society and its board in the late 1980s, when he cochaired the Social and Demographic Committee.

In 1996, Charles was elected a fellow of ASA “for excellence in the application of statistical methodology in the legal and regulatory communities; and for leadership in developing awareness and consideration of professional ethics in the practice of statistics.”

Charles held a BS in applied mathematics from Brooklyn’s Polytechnic Institute, an MS in mathematical statistics from Michigan State, and a PhD in statistics from the University of Missouri. He was an instructor at the University of Maine, Orono, and an assistant professor in the department of statistics at The George Washington University. Before he opened his own firm, he headed and became a vice president of the statistical services division of Group Operations, Inc., in Washington, DC. He is the author of the entry on underutilization in the Encyclopedia of Statistical Sciences.

In 1977, Charles became founder and president of Charles R Mann Associates and was one of the pioneers in the provision of statistical services to legal practitioners in employment law. He participated in more than 100 seminars, workshops, and individual presentations to lawyers on various statistical applications and methodologies to be applied for proper statistical evaluation of employment discrimination claims. He also helped educate the legal profession to use multiple regression analysis to evaluate equal pay claims and use only statistically significant differences in application and/or employment rates for establishing a prima facie case. In addition to providing lawyers with advice and reports on statistical analyses, Charles provided expert witness testimony in more than 160 cases, including major and landmark cases in a variety of legal arenas, and participated personally in settlement negotiations for several major national class action cases.

The Equal Employment Opportunity Act of 1972 first authorized the EEOC and private individuals to bring lawsuits under Title VII, which prohibits employment discrimination on the basis of race, sex, and national origin. In 1973, a nationwide class action lawsuit—O’Bannon and EEOC v. Merrill Lynch, Pierce, Fenner & Smith, Inc.—alleged Merrill Lynch discriminated against women applicants for a stockbroker position while the EEOC alleged discrimination on the basis of race, sex, and national origin in all positions. Charles helped forge a defense for the Merrill Lynch lawyers, which resulted in an affirmative action consent decree that adopted Charles’s unique statistical methodology for establishing goals and timetables for hiring minorities and women. This ‘five year plan’ compares the percentage of minorities and women actually employed with the percentage in the ‘comparable’ workforce and then establishes goals and timetables when a statistically significant difference exists in the employment rate. This was a landmark settlement and set the standard for the brokerage industry.

Many early employment law cases were litigated by parties looking only at the number and percentages of minorities and women actually employed, versus their corresponding number and percentages in the civilian labor force. This was commonly referred to as the “any difference rule.” For example, where a company had no women employed out of five electrical engineers compared to 20 percent of women in the labor force, a prima facie case of discrimination was established under the “any difference” rule. Charles was tireless in his attack on such uncritical thinking and incorrect methodology. He urged lawyers for both plaintiffs and defendants to focus on the “expected” numbers: to use statistics in a more refined way to ascertain whether the null hypothesis of nondiscrimination should be rejected.

Charles was also involved in a variety of discrimination cases outside employment law, such as Anne Brent and the Fair Housing Center of Metropolitan Detroit v. Henry Ford Village, et al., a 1997 case alleging discrimination under the Fair Housing Act in the methods of advertising. There was also Kernan v. Holiday Universal, United States District Court for the District of Maryland, a defense of a nationwide class action lawsuit alleging racial discrimination under the federal public accommodation statutes involving health club memberships.

Parallel with his work in employment discrimination litigation, Charles provided expertise in the original development of affirmative action programs for federal contractors and subcontractors. This longstanding Department of Labor program was established by Executive Order 11246 and administered by the Office of Federal Contract Compliance Programs (OFCCP). Among other regulations, the order requires government contractors and subcontractors to “… take affirmative action to [ensure] that applicants are employed, and that employee[s] are treated during employment, without regard to their race, color, religion, sex, or national origin.” As enforcement increased, the secretary of labor approved an interpretation that authorized a rule requiring contractors to seek to achieve and maintain a percentage of women and minority workers exactly equal to their estimated availability in the external civilian workforce. Charles was first to recognize this as another version of the “any difference” rule in litigation and vigorously opposed it, waging nearly a one ‘Mann’ war with the OFCCP to recognize statistical procedures to measure more accurately whether the differences were statistically significant.

The OFCCP then abandoned the “any difference” rule and adopted the 80 percent rule, which Charles likewise fought. That is, the percentage of minorities and women in the internal workforce had to be at least 80 percent of their corresponding availability in the external workforce. As Charles successfully educated the legal profession in using proper statistical analyses, it became clear there needed to be a sound statistical basis for taking affirmative action in setting goals and timetables for the employer to avoid direct violations of Title VII. This dispute led to the seminal case that vindicated Charles’s views and can be seen as the crowning achievement in his career of advising the courts and legal profession in statistics in the employment context. Firestone Synthetic Rubber & Latex Co. & Koppers Company, Inc. v F. Ray Marsha, et al., 507 F. Supp. 1330 (E.D. Tex. 1981) was a declaratory judgment action brought to prevent the debarment of a contractor for, among other things, only declaring underutilization of minorities and women when their internal employment rate was statistically significantly less than their availability in the corresponding labor force (at the .05 level). Enforcement of the “any difference” rule was finally ended.

In addition, Charles’s use of regression to analyze pay data was the first to be accepted by the Supreme Court in Bazemore v. Friday. That decision established the criteria courts now use to evaluate the reliability of regression analyses submitted in litigation and to justify affirmative action plans.

Charles was a highly esteemed mathematical statistician, responsible for the court’s understanding and adopting statistical analyses and methodologies still recognized and used in a wide variety of cases today, benefiting lives by helping to establish rigorous and transparent statistical methods for accountability in the legal system.

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)

Comments are closed.