Simulations? [Power / Sample Size]

posted by mittyri – Russia, 2019-01-31 23:28  – Posting: # 19836
Views: 1,116

Hi,

if you don't have analytical solution - look for simulations approach :-D
Simulate the data using known distribution parameters, desired sample size and then use bootstrap technique for evaluation. Repeat it ad-hoc (until the success/failure ratio stabilizes). Then change the sample size and repeat until desired power is reached. Don't forget about SciAdv.

Kind regards,
Mittyri

Complete thread:

Activity
 Admin contact
20,817 posts in 4,356 threads, 1,447 registered users;
online 33 (0 registered, 33 guests [including 16 identified bots]).
Forum time: 10:28 CEST (Europe/Vienna)

Statistics are no substitute for judgment.    Henry Clay

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