William J. Studden passed away suddenly on March 19th, 2013.
Bill was born in Hamilton, Ontario, Canada on September 30, 1935. Professor Studden received his B.Sc. from McMaster University (1958) and his PhD in Statistics in 1962 from Stanford University for a thesis on Asymptotic Laws for Birth and Death Processes under the supervision of James McGregor and Samuel Karlin. At the same time he was working intensively with Samuel Karlin writing the book on Chebyshev Systems. He joined the Department of Statistics at Purdue University in 1964. Although Bill retired in 2005 as a faculty member in Statistics, he remained active in research, was an avid reader, enjoyed exploring genealogy, and had a lifelong interest in antiques.
Bill Studden served as Advisory Editor of Journal of Statistical Planning and Inference for many years and he was a member of Sigma Xi, a Fellow of the Institute of Mathematical Statistics, and an elected member of the International Statistics Institute.
Bill was a leading figure in the field of optimal experimental design. During his career Bill published over 80 articles in peer-reviewed journals; he co-authored two books. His publications cover a broad range in mathematics, statistics and probability. He is very well known beyond the statistics community for his book with Samuel Karlin on Chebyshev Systems. In statistics Bill has made important contributions in the field of optimal experimental design. At the beginning of his career Bill worked on geometric characterizations of optimal designs in regression models and provided a deeper understanding of the famous Elfving Theorem. He also worked with Jack Kiefer on designs for large degree polynomial regression models. In the middle of the eighties he used classic results in mathematics on continued fractions and orthogonal polynomials to develop the theory of canonical moments of probability measures on a compact interval. These are in one-to-one correspondence with the common moments but provide an extremely useful and powerful tool for solving difficult design problems in polynomial regression models explicitly. Interestingly, these objects are related to Verblunsky coefficients, which have recently found much interest in mathematical physics and random matrices.
Bill advised sixteen PhD students. His office door was always open and he generously shared his great mathematical ideas with his colleagues. Bill is survived by his wife, Myrna M. Harrison Studden and numerous friends in the statistics and mathematics world. He will be remembered as an excellent scientist in mathematical statistics working on very deep problems until he had understood and solved them completely.
Written by Holger Dette,
Ruhr-Universität Bochum, Germany