ABSTRACT
From a logistical and cost-effectiveness point of view it is natural to employ sequential statistical procedures. The sequential analysis methodology possesses many advantages, including economic savings in sample size, time and cost. It also has advantages in terms of ethical considerations and monitoring, e.g. a clinical trial might be stopped early if a new treatment is more efficacious than originally assumed. There are formal guidelines for using of sequential statistical procedures in many research areas. Sequential methods are well suited for use in clinical trials with short term outcomes, e.g. a one-month follow up value. When dealing with human subjects, regular examinations of accumulating results and early termination of the study are ethically desirable. This chapter outlines several of the more well-applied sequential testing procedures, including two-stage designs, the sequential probability ratio test, group sequential tests, and adaptive sequential designs. Chapter 6 introduces a technique that can be applied to evaluate asymptotic properties of stopping times used in statistical sequential schemes. Corresponding R codes are demonstrated. Chapter 6 indicates critical problems related to complicated estimation and testing issues in the post-sequential analyses of the data.
