Recent developments in multivariate meta-analysis and a robust approach based on the theory of U-statistic

  • Created: 2012-04-17
Time: 10:30-11:30 am, April 20, 2012

Speaker:Yan Ma
Venue:Teaching Building Room 709, Yuquan Road
Dr.  Ma is an Assistant Professor of Biostatistics in the Division of Biostatistics and Epidemiology in the Department of Public Health at Weill Medical College of Cornell University and Assistant Professor of Biostatistics in the Research Division at the Hospital for Special Surgery. He received his Ph.D. in Statistics from the University of Rochester in 2008. His research interests are statistical methodology center on inference for instrument reliability and rater agreement, nonparametric estimation and inference, observational studies, causal inferences, meta-analysis, complex survey data analysis, longitudinal data analysis, and missing data problems.
Abstract: With rising cost of medical research, many clinical trials are carried out with small sample sizes resulting in low power for detecting clinically useful effect size. By pooling the estimated effect sizes from each of the component studies through meta-analytic methods, information from larger number of patients and increased power is expected. Because of its important impact, development of novel methods for meta-analysis has become a rapidly growing field. In this talk, I will review the literature with emphasis on recent developments in multivariate meta-analysis (MMA) and present Ma and Mazumdar (Stat in Med, 2011)’s new method for MMA.