Frank Miller: Applied sequential designs and statistical issues
Tid: On 2013-11-13 kl 13.00 - 14.00
Plats: Room B705, Department of statistics, Stockholm university
Medverkande: Frank Miller, Department of statistics, Stockholm university
Sequential designs have been applied in several areas including clinical trials and recently for online controlled experiments to improve web-page layout. The talk introduces specifically the design types “group sequential design” and “sample size re-estimation” and how they are applied. We discuss how these designs impact statistical inference: How to do significance testing, estimation and computation of confidence intervals for these designs?