ISI - International Statistical Institute Newsletter Volume 26, No. 3 (78) 2002
In Memoriam
|
Deceased Members |
|||
|
The ISI regrets to announce the death of our colleagues: |
|||
| Born | Elected | Deceased | |
| Prof. Jerzy Z. Holzer | 1930 | 1974 | 1 September 2001 |
| Prof. Justus F. Seely | 1941 | 1983 | 23 February 2002 |
| Prof. Kali S. Banerjee | 1914 | 1973 | 9 April 2002 |
| Prof. Nathan Mantel | 1919 | 1971 | 25 May 2002 |
| Prof. George A. Bernard | 1915 | 1952 | 31 July 2002 |
| Prof. Silvio Vianelli | 1910 | 1957 | 2002 |
| Prof. Pierre Thionet | 1915 | 1955 | 2002 |
 |
Kali S. Banerjee (1914-2002)
Professor Kali S. Banerjee, a veteran and distinguished Indian statistician and ISI member, died on Tuesday April 9, 2002 at his Long Island, NY residence, aged 88. Born in 1914, Prof. Banerjee received his training in mathematics and statistics from the University of Calcutta, earning his PhD. In statistics in 1950. His fundamental work on weighing designs in 1948 soon brought him international acclaim. |
After serving as Deputy Director at the West Bengal State Statistical Bureau in West Bengal, India from 1951 to 1962, Prof. Banerjee briefly visited Cornell University (1962-1963) and Kansas State University (1964-1966, 1968-1969), and then accepted a permanent position at the University of Delaware (1969-1979), holding the Rodney Sharp Chair Professorship. After retiring from UD, Prof. Banerjee joined the University of Maryland Baltimore County (UMBC) as a Visiting Professor, a position he held until his eventual retirement from academic work in 1986.
Prof. Banerjee’s original contributions in the areas of general linear models, weighing designs, factorial designs, and theory of index numbers, the areas in which he published widely, continue to have a profound impact in the statistics literature.
As well as being an ISI member, he was a Fellow of the American Statistical Association, the Royal Statistical Society, the American Association for the Advancement of Science, and the IMS. Prof. Banerjee will be missed by his colleagues and admirers all over the world.
Bimal K. Sinha
Justus Frandsen Seely (1941- 2002) Justus Frandsen Seely, Professor of Statistics at Oregon State University and an influential researcher in linear models, died on February 23, 2002 of an apparent heart attack while attending a high-school girls' basketball game coached by one of his daughters.
Justus was born on February 11, 1941 in the small town of Mt. Pleasant, Utah. He grew up there and spent several summers as a youth working on his father's sheep ranching operation. He attended Brigham Young University for two years and then transferred to Utah State University, where he received a BS degree in statistics in 1963 and an MS in statistics in 1965. Justus received his doctoral training in statistics from Iowa State University, where he received the Snedecor Award as the outstanding student in statistics.
After completing his doctoral degree in 1969, ustus joined the Department of Statistics at Oregon State University in Corvallis, Oregon. He served as Chair of the department from 1986 to 1997 and as Associate Dean of the College of Science from 1997 until his death. He was a fellow of the American Statistical Association and the Institute for Mathematical Statistics, an elected member of ISI, and the recipient of four awards for excellence in teaching.
Justus made various insightful contributions to the theory of Gaussian linear models. In his Ph.D. research he pioneered the concept of extending linear estimation theory to explicitly include quadratic statistics for estimation of variance components. His adviser, George Zyskind, had elegantly characterised the situation where ordinary least squares estimation of fixed effects remains optimal for mixed models: the regression space should be invariant under multiplication by the covariance matrix. Justus extended this idea to variance component estimation, introducing the notion of invariant quadratic subspaces and their relation to completeness. By characterising the class of admissible unbiased estimators of variance components, he showed that the usual ANOVA estimator in 2-variance-component models is inadmissible. Among other contributions to the theory of mixed models, he succeeded in generalising and improving on several existing procedures for tests and confidence intervals on variance components.
Justus never strayed far from his rural roots. His favourite hobbies were gardening and fishing. Students studying on weekends and evenings could often find Justus in his office wearing his preferred attire, a pair of overalls. His easy approachability and concern for students made him a popular mentor, and he produced a steady stream of Ph.D. students in linear models. Justus always had time for those who needed help, and he was unfailingly fair and generous with students and colleagues alike.
Justus is survived by his wife, Averil, three children, and five grandchildren.
Dave Butler
Back to Home
Back to Newsletter
Top