Delaware Chapter, ASA
Dinner Meeting: September 2008
Note: New location just for this September meeting:
The Statistics Program at the University of Delaware, College of Agriculture and Natural Resources is hosting a special Kick-off meeting of our DE ASA program in the Commons of Townsend Hall at 6 pm on Thursday, September 18, 2008. Please come early to enjoy UD's Botanic Gardens from 5 to 6 pm (http://ag.udel.edu/udbg).
The Happy Hour will begin at 6 pm and dinner around 6:45. Our speaker will be the distinguished Professor Sanat K. Sarkar from Temple University who will speak on "Techniques of Modern Multiple Testing." Reservations are not necessary, but greatly appreciated.
Townsend Hall is located on S. College Ave. in Newark on the southern campus, next to the ice arenas and Field House. The Commons area is located in the center part of the building. Directions can be found at:
Call Tom Ilvento 302-388-0971 if your need more directions.
A flyer is attached, please post it in your work area.
Speaker: Sanat K. Sarkar
Fox School of Business and Management
Title: Techniques of Modern Multiple Testing
In the past decade, there has been renewed interest in multiple testing due to its increased relevance as a statistical tool to analyze data from modern scientific investigations, such as genomics and brain imaging. As these experiments typically involve a large number of null hypotheses to test, making some of the more traditional multiple testing methods, like those designed to control the familywise error rate (FWER), not as powerful as one would like, the research in modern multiple testing has been focused to a large extent on developing more meaningful and powerful notions of error rate than the FWER and methods that control them. The false discovery rate (FDR) introduced by Benjamini and Hochberg (1995) is one of these error rates that has been the most popular so far. In this talk, I will present some recent developments of methods that are designed to control the FDR.
My research interests in statistical theory span many general areas: Statistical inference, multivariate statistics, probability inequalities, categorical data analysis, pharmaceutical statistics and multiple testing. I worked for my Ph.D. in the area of multivariate statistical inference with missing data. Later I became interested in inequalities which gradually evolved into research in other areas. In particular, I established power properties of multivariate tests, developed confidence regions in time series, provided meaningful models for bivariate survival data and ordinal contingency tables, I have published several papers on multivariate estimation including some involving shrinkage methodology. I did research on statistical inferences that arise in drug development process. These include testing of hypotheses in analysis of combination drugs and bioequivalence studies, estimation of parameters under order restrictions. In addition, I have worked on problems involving order statistics from dependent random variables. Recently, I've published a number of papers in the area of multiple testing.
I am fellows of the American Statistical Association and the Instute of Mathematical Statistics, and an elected member of the International Statistical Institute.
Date: Thursday, September 18, 2008
Time: Social hour -- 6:00 pm
Dinner -- 6:30 pm
Speaker -- 7:15 pm
Place: Townsend Hall
South College Ave.
University of Delaware
Cost: $25 for members, $15 for students
Irina Baldytcheva by Wednesday, September 17, 2008. Phone: 302-885-9136 Email: Irina Baldytcheva.