Qing Pan and Doug Schaubel, Lauréats du prix de La revue canadienne de statistique 2009
The Statistical Society of Canada is pleased to award Qing Pan and Doug Schaubel the Canadian Journal of Statistics Award for the best paper published in the journal in 2008. The papers are judged according to excellence, innovation and presentation.
The paper, titled "Proportional hazards models based on biased samples and estimated selection probabilities," was published in Vol. 36, No. 1, 2008, pp. 111-127.
Unrepresentative samples are common in observational studies and often lead to biased parameter estimates. The authors propose a two-stage inverse-probability-of-selection weighted proportional hazards model, using weights estimated from auxiliary information on the sampling process. The estimation of the weights is explicitly incorporated into the inference procedures, which leads to gains in efficiency relative to existing methods that treat the weights as fixed. The method is widely applicable from epidemiologic to ecological studies. Through the proposed methods, Pan and Schaubel demonstrate that the increased failure risk associated with expanded criteria on donor kidneys is greatly underestimated by previous analyses which did not account for the inherent bias introduced by the acceptance/discard process.
Dr. Pan is an Assistant Professor in the Department of Statistics at George Washington University. A winner of ENAR's 2008 Distinguished Student Paper Award, her research interests center on survival analysis, recurrent event data, as well as their applications in observational studies, clinical trials and equal employment cases. She obtained her Ph.D. from University of Michigan. An Associate Professor in the Department of Biostatistics at the University of Michigan, Dr. Schaubel's research interests include multivariate survival analysis, recurrent event data, dependent censoring, and epidemiologic studies. He holds B.Sc., M.Sc. and Ph.D. degrees from University of Waterloo, McGill University and University of North Carolina, respectively.