Richard Lockhart, SSC Gold Medalist 2015
The 2015 recipient of the Gold Medal of the Statistical Society of Canada is Professor Richard Lockhart. The Gold Medal is awarded to a person who has made outstanding contributions to statistics, or to probability, either to mathematical developments or in applied work.
Richard Lockhart is Professor in the Department of Statistics and Actuarial Science at Simon Fraser University. Richard was born in Montréal in 1954 and grew up in Winnipeg, Manitoba and Tsawwassen, British Columbia. He received a BSc in Math from UBC in 1975, an MA from the University of California at Berkeley in 1976 and then a PhD in Statistics from the same university in 1979. His thesis "The Programming Operation on s-fields" was written under the supervision of David Blackwell. After a six-month post-doctoral fellowship at the Centre de recherches mathématiques at l'Université de Montréal he joined Simon Fraser University in September 1979. He has been at SFU since then except for sabbaticals at Waterloo and Oxford and a single year at the University of Toronto.
Richard is a statistical leader in Canada, through his research, his many contributions to the Statistical Society of Canada, his teaching and mentoring of students at Simon Fraser University, but most of all through his enthusiasm for good ideas and the pleasure of tackling new problems. He has had a long-standing interest in goodness-of-fit testing and, with Michael Stephens, has made a number of important contributions, from their Biometrika paper in 1985 through to his Bernoulli paper in 2012. This last paper was an elegantly written, thoughtful and very original contribution to a topic that many people have dismissed as 'finished'. Richard showed that conditional and unconditional tests of goodness-of-fit may be expected to be nearly identical under certain conditions. The importance of this contribution will take some years to become clear; the set of remarks in the concluding discussion section outline a wealth of ideas and a number of ways the work could be investigated further.
Although known for his work in goodness-of-fit, Richard is one of our discipline's few polymaths - he has made interesting and original contributions to a surprising variety of topics. With his Berkeley roommate, Peter Guttorp, he worked on Bayesian models for directional data, uniform limit theory for high dimensional quadratic forms and inference in Galton-Watson-Bienaymé processes. With his former student, Grace Chiu, he has written several papers on so-called 'bent-cable' regression. He has had an interesting collaboration with Peter Borwein investigating random polynomials; their papers appeared in the Annals of Mathematics and the Proceedings of the American Mathematical Society. He collaborated with Joan Hu in work on panel data, and with his former student, Gemai Chen, on the Box-Cox model. His collaborators would confirm that his contributions and his insight were essential to the success of the project. In the past few years he became involved in a very exciting collaboration with Rob and Ryan Tibshirani and Jonathan Taylor on asymptotic theory for lasso estimators. This is a very important topic that is only just being addressed by a handful of people working in high-dimensional data. That Richard was able to make important contributions without having an extensive background in the field is very impressive, although knowing his strengths, not surprising.
Richard's great strength is a deep knowledge of analysis, probability, and theoretical statistics that has enabled him to solve problems in asymptotic distribution theory and rigorous justification of procedures. He has always been ready to tackle any problem brought to him and this is shown by a wide range of publications and collaborators, with major contributions to goodness-of-fit, signal processing, stochastic processes, use of the Box-Cox transformation, smoothing, and the lasso, among others.
Richard's contributions to professional service are equally noteworthy and wide ranging, and many of these relate to research. He served as Editor of the Canadian Journal of Statistics (2001-03), and as the Statistical Society of Canada's Executive Editor for Statistical Surveys (since 2007); he was also an Associate Editor of Technometrics from 2002-07. He has served as President of the Statistical Society of Canada (1996-97), and on numerous scientific advisory committees: for example, as a member of Statistics Canada's Advisory Committee on Statistical Methods (1998-2012), and as a member of the ASA Advisory Committee to the Energy Information Agency (1991-96). He was a member of the Natural Sciences and Engineering Research Council of Canada's grant selection committee for statistical science (1991-94), chairing a joint Mathematics and Statistics equipment grant selection committee in his final year. He has been the Program Chair for the Statistical Society of Canada's Annual Meeting (2006), and has served on several other organizing committees. From 2008 to 2014, he served as Chair of the Department of Statistics and Actuarial Science at Simon Fraser University.
Richard's contributions have been recognized before. He was elected Fellow of the American Statistical Association in 2013 and a Fellow of the Institute of Mathematical Statistics in 2014. He received the Distinguished Service Award of the SSC in 2002.
One letter writer said "Richard is the most unselfish researcher I have ever known - he enjoys every opportunity to offer his help, regardless whether the question came from his students, from his colleagues, from a paper under review, from consulting, or from an email originated far away, and he often takes no credit."
The citation for the award reads:
"To Richard Lockhart, for outstanding contributions to statistical inference and methodology; for development of asymptotic distribution theory and rigorous justification of procedures in applied statistics through his deep knowledge of analysis, probability, and theoretical statistics; for the breadth of his contributions notably on goodness-of-fit, signal processing, stochastic processes, use of the Box-Cox transformation, smoothing, and the lasso, among others."