Sample size justification [Design Issues]

posted by Helmut Homepage – Vienna, Austria, 2015-08-07 03:21 (3513 d 22:57 ago) – Posting: # 15189
Views: 11,996

Hi Balaga,

❝ "what is the maximum expected sample size (for two way crossover design) with the intra subject coefficient of variation of pilot study data around 15%.


It depends. ;-)

❝ I have the following data,

❝ Observed Pilot study ISCV% - 12.29%


Sample size of the pilot?

❝ With the observed pilot study (two way crossover design) ISCV 12.29%, is it possible to expect the 15% ISCV or not for pivotal study design?


Yes. Once you’ll tell us the sample size of the pilot maybe more about it…

❝ What should be the null ratio considerate for the same and under what circumstances we should go for widening of the null ration beyond the 95 - 105% limits or we can in all cases?


If you are not dealing with a HVD/HVDP and the GMR in the pilot was within this range, fine. Otherwise expecting a larger deviation is not a bad idea.

❝ During sample size calculation, is there is any restriction to expect the null test/reference ratios in between 95.00 to 105.00% only? I wont feel so.


Correct. BTW, power curves are symmetric around 0 in the log-domain. Therefore, you get the same power at 95% and at 1/95% ~105.26% (log ±0.051293). If you expect a 5% deviation of T from R (and are not sure about the direction), always estimate the sample size at 95%.

❝ […] if there is any separate link to access the Diletti et. al. (1991) sample size calculation article.?


Check your [image].
IMHO, Diletti’s “Table 1” is of historic interest only (too wide step sizes).

Dif-tor heh smusma 🖖🏼 Довге життя Україна! [image]
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

Complete thread:

UA Flag
Activity
 Admin contact
23,409 posts in 4,921 threads, 1,670 registered users;
15 visitors (0 registered, 15 guests [including 3 identified bots]).
Forum time: 01:18 CET (Europe/Vienna)

Ignorance more frequently begets confidence
than does knowledge.    Charles Darwin

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