An approach theoretic version of Anscombe's theorem with an application in biostatistics

Abstract

We establish an approach theoretic version of Anscombe's theorem, which we apply to justify the use of confidence intervals based on the sample mean after a group sequential trial.

arXiv comment7 pages