PowerTOST? [Power / Sample Size]

posted by Helmut Homepage – Vienna, Austria, 2016-04-18 15:22 (3249 d 15:03 ago) – Posting: # 16211
Views: 7,251

Hi Junhao,

you can try to get it from the the Publishers (Dustri-Verlag, contact Petra Estner). Alternatively you could ask Dieter Hauschke for a copy.
If that doesn’t work I can scan my copy (will be lousy: yellowed and with notes in the margins…).
It is a very interesting paper if you want to understand the background of sample size estimation. For practical use less so. The tables are for 70/80/90% power, T/R-ratios of 0.85–1.20 (step size 0.05), and CVs of 5–30% (step size 2.5%). Power curves are highly nonlinear. If you want anything which is not covered in the tables, interpolation is not a good idea.
If you want to estimate sample sizes for any desired power, ratio, and CV I recommend the package PowerTOST for R (freeware!). In PowerTOST many designs are covered (two parallel groups, paired means, crossovers: 2×2×2, 3×3, 3×6×3, replicates: 2×2×4, 2×2×3, 2×4×4, partial 2×3×3, Balaam’s 2×4×2, Liu’s 2×2×2). Additionally to power / sample size for conventional average bioequivalence other methods are supported: Reference-scaling for HVD(P)s according to the FDA’s, EMA’s, and ANVISA’s requirements; RSABE for NTIDs (FDA); noninferiority, ratio of means (Fieller’s CI), dose proportionality (power model). Check out the online-manual.

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
23,406 posts in 4,919 threads, 1,744 registered users;
44 visitors (0 registered, 44 guests [including 8 identified bots]).
Forum time: 05:26 CET (Europe/Vienna)

“Data! Data! Data!” he cried impatiently.
“I can’t make bricks without clay!”    Arthur Conan Doyle

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