Activity Number:
|
295
- Adaptive Designs and Interim Analyses
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, August 1, 2017 : 8:30 AM to 10:20 AM
|
Sponsor:
|
Biopharmaceutical Section
|
Abstract #322528
|
View Presentation
|
Title:
|
A Curtailed Procedure for Comparing Treatments with Binary Response with a Control
|
Author(s):
|
Pinyuen Chen* and Lifang Hsu
|
Companies:
|
Syracuse University and Le Moyne College
|
Keywords:
|
Indifference Zone Approach ;
Subset Selection Approach ;
Integrated Formulation ;
Curtailment ;
Least Favorable Configuration ;
Expected Sample Size
|
Abstract:
|
We propose a sequential procedure for comparing the success probabilities of k (>1) experimental Bernoulli populations among themselves and with the success probability of a controlled Bernoulli population. Our approach follows the integrated formulation considered by (1) Chen and Sobel (1987) whose goal was to select the best among k normal means and by (2) Hsu (1991) who compared k experimental Bernoulli populations with a controlled Bernoulli population. The propose procedure takes vector observations one at a time with possible early termination. Using a smaller expected sample size, the procedure satisfies the same two probability requirements as does the fixed sample size procedure: (i) the probability of correctly selecting the best treatment is at least p1 when the best is significantly superior to the others and (ii) the probability of including the best treatment in the selected subset is at least p2 when the difference between the best and the others is not significant.
|
Authors who are presenting talks have a * after their name.