Noise… [RSABE / ABEL]

posted by Helmut Homepage – Vienna, Austria, 2014-07-22 15:31 (3948 d 22:20 ago) – Posting: # 13290
Views: 13,027

Dear Detlew,

❝ Where does the tremendous noise in your pictures came from?


Quick and dirty as ever; I set a different seed for every simulation. Looks better if keeping the same one:

[image]

[image]

❝ And Fig. 2 (as well as others) of the paper is with tremendous (arbitrary?) smoothing? Crossing of some of the curves seems to point in that direction. The plot form (continuous curves, without the points they have simulated) itself is suspicious IMHO.


Agree.

❝ BTW: Do you see any need to create a power.addon() in PowerTOST? Pedagogical?


I guess enthusiasts could already misuse* the – experimental – function power.2stage.GS() as I did. I tried alpha=c(0.05, 0.05) and n=c(n, n/2+(n/2)%%2). The modulo in the second part of the n-vector rounds up to the next even number be­cause I don’t like imbalanced studies. Example: n = 18, n/2 = 9, n/2+(n/2) mod 2 = 10.
power.addon() would be nice, of course. ;-)

BTW, the Japanese guidance states:

The add-on subject study should include at least one half of the number of subjects in the initial study.

I overlooked that until today (believing that n2 is fixed at n1/2). This opens the door to an infinite number of designs (adjusting for the observed CV, any power, even fully ad­ap­tive). Glad that I never dealt with submissions to Japan so far.



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
23,424 posts in 4,927 threads, 1,672 registered users;
154 visitors (0 registered, 154 guests [including 8 identified bots]).
Forum time: 13:52 CEST (Europe/Vienna)

Only dead fish go with the current.    Scuba divers' proverb

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