Dr. Xianming Tan, research associate professor of biostatistics at the UNC Gillings School of Global Public Health, has received the Statistical Society of Canada’s Canadian Journal of Statistics Award for his paper, “Variable selection and inference procedures for marginal analysis of longitudinal data with missing observations and covariate measurement error,” published online in the journal October 20, 2015.
[Photo: Dr. Xianming Tan]
The annual award, which recognizes a paper’s outstanding methodological innovation and presentation, was presented at the annual meeting of the Statistical Society of Canada, held May 29 – June 1 at Brock University in St. Catharines, Ontario.
Research on model selection for longitudinal data remains largely unexplored, especially when data is missing and subject to measurement error. The paper proposes marginal methods that simultaneously carry out model selection and estimation, while allowing for missing responses and error-prone covariates.
The awarding committee praised the paper for its combination of applicability, strong theoretical work and solid practical assessment of the proposed methods.
“This research is important because it tries to approximate the real complexity of longitudinal data to a great extent, and proposes a practical and theoretically sound approach to handle such complexities involving measurement error, drop-out, and model selection, simultaneously,” said Dr. Tan. “It might be the first published work on this topic in longitudinal data analysis and will motivate new research as well as applications of the proposed approach in longitudinal studies.”
Dr. Grace Y. Yi, professor of statistics and University Research Chair at the University of Waterloo in Ontario, Canada, and Dr. Runze Li, Verne M. Willaman Professor of statistics at The Pennsylvania State University, are co-authors of the winning paper.