Bachhofer-Gupta-Sobel Memorial Symposium on
Ranking and Selection and Multiple Comparison Procedures Methodologies
|
Sanat K.
Sarkar
is Professor of Statistics and Senior Research Fellow in the Fox
School of Business at Temple University. He has published several
articles in a wide variety of statistical areas, such as,
categorical data analysis, multiple testing, multivariate
statistics, pharmaceutical statistics, probability inequalities
and survival analysis, many of which have appeared in top-tier
journals. Recently, he has made ground-breaking contributions to
the area of multiple testing, having solved a conjecture that had
been outstanding for more than a decade. He co-edited two volumes
and presented invited talks at numerous national and international
conferences. His research has been funded by NSF and NSA. He has
supervised fifteen Ph.D. students. He was named to the research
honor roll three times and awarded the Musser Award for excellence
in research by the Fox School of Business. He is a Fellow of the
American Statistical Association, an elected member of the
International Statistical Institute. |
Generalizing Holm’s and Hochberg’s
Stepwise Procedures for Multiple Hypothesis Testing
Having realized that the familywise
error rate(FWER) is too stringent to control when large number of
hypotheses are tested, researchers have recently focused on defining
less stringent error rates and developing methods that control them.
Recently, Lehmann and Romano (2005, Annals of Statistics) and Romano
and Shaikh (2006, to appear in Annals of Statistics) have considered
controlling the k- FWER, the probability of falsely rejecting at least
k null hypotheses, and proposed single-step and stepwise procedures
that control this error rate. Newer and often more powerful procedures
controlling the k-FWER will be presented in this talk. These
procedures are developed based on a generalization of Simes’ test.
|