|
|
|
|
37th Annual Craig Lectures
October 11 and 12, 2007
Nancy Reid
Department of Statistics
University of Toronto
 Nancy Reid is Professor of Statistics at the University of Toronto. She received her BMath from the University of Waterloo, her MSc from the University of British Columbia, and her PhD from Stanford University. She held a faculty appointment at the University of British Columbia before joining the Department of Statistics at the University of Toronto in 1986.
Nancy is a leading statistical scientist whose work has had major impact in the area of theoretical statistics. Her work on the asymptotic theory of likelihood has led to new advances in understanding of the basis of statistical inference as well as new techniques for application. Her current interests also include statistical methods in biology and environmental science. She was the first woman and first resident Canadian to receive the Presidents' Award of the Committee of Presidents of Statistical Societies. She has served as President of the Statistical Society of Canada, President of the Institute of Mathematical Statistics, and Vice-President of the Internationa Statistical Institute, and is Fellow of the Institute of Mathematical Statistics, the American Statistical Association, the Royal Society of Canada and the American Association for the Advancement of Science.
|
|
Lecture #1
Thursday, October 11
Refreshments at 3:00pm in 241 SH
Lecture at 3:30pm in 140 SH
"Weighting the Likelihood Function"
Statistical theory is often categorized as either "Bayesian" or "frequentist", and statisticians often self-identify in the same categories. With the development of several key results in the asymptotic theory of inference based on the likelihood function, it is becoming clear that the statistical differences between Bayesian and frequentist methods are rather less important than the philosophical ones. One aspect of this is an ongoing effort to develop so-called 'reference', or 'objective' or 'default' priors. I will give an overview of some of the asymptotic theory behind the development of approaches to constructing such priors, with special attention to strong matching priors that have been developed recently in joint work with Don Fraser and colleagues. The construction of these priors provides some insight into the exact points of departure between Bayesian and frequentist methods, at least from the statistical point of view. The philosophical debate may well continue for some time.
|
Lecture #2
Friday, October 12
Craig Cake and Punch at 2:30pm in 241 SH
Lecture at 3:30pm in 140 SH
"Putting Asymptotics to Work"
Asymptotic theory based on the likelihood function has been extensively developed in the theoretical literature over the past twenty-some years, and this has led to a number of so-called 'higher order approximations' that often give surprisingly accurate results. In addition to providing excellent approximations, the theory provides insight into the process of statistical inference. I will give a non-technical overview of the main developments, by illustrating the use of the results in various types of regression models for continuous and discrete data. This is based on joint work with Alesssandra Brazzale and Anthony Davison.
|
|
