Confuse-a-Cat [Two-Stage / GS Designs]

posted by ElMaestro  – Denmark, 2017-08-19 19:56 (3219 d 07:19 ago) – Posting: # 17715
Views: 33,898

Hi again,

❝ Now I get it – though your code is beyond me. Zhang’s method in many cases hits the right point in the first attempt. Sometimes sampleN.TOST() has to iterate upwards (since Zhang’s method is based on a large sample approximation and power with anything of the t-family will be lower). In very rare cases sampleN.TOST() has to iterate downwards. Can your code do that as well?


Yes. Basically, get the initial initial guess, determine if it is low or high for the desired power, iterate up or down depending on that.

The reason I am doing this is I got a mental "Does not compute error" when I read Zhang's paper. The goal is to get to the right sample size in as few cycles as possible.

Pass or fail!
ElMaestro

Complete thread:

UA Flag
Activity
 Admin contact
23,653 posts in 4,991 threads, 1,571 registered users;
153 visitors (0 registered, 153 guests [including 40 identified bots]).
Forum time: 03:16 CEST (Europe/Vienna)

To propose that poor design can be corrected by subtle analysis techniques
is contrary to good scientific thinking.    Stuart J. Pocock

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5