SAS/R code of MSD method for comparing dis­so­lution [Dissolution / BCS / IVIVC]

posted by Shuanghe  – Spain, 2016-12-22 10:23 (3119 d 22:01 ago) – Posting: # 16873
Views: 16,022

Hi,

❝ OK, I now know why these values are different.

❝ DDsolver basically calculates the lower and upper bounds as follows:

❝

❝

❝ df    <- nt + nr - p - 1

❝ k     <- nt * nr / (nt + nr) * df / ((nt + nr - 2) * p)

❝ fcrit <- qf(0.9, p, df)

❝

❝ dm_lower <- dm - sqrt(fcrit/k)

❝ dm_upper <- dm + sqrt(fcrit/k)

❝

❝

❝ That is quite a bit simpler, is it also correct?


DDSolver's source is locked, so I thought that, unless authors let you, no one can have a peek as how it works exactly. I have no idea how you figure this out but :thumb up: from me.

If they obtain low and upper limits as you said then I don't think it's correct. I'm not a statistician so my thinking might be naïve. For me, here the Mahalanobis distance is a measure of the difference in terms of "distance" and when there is no difference the distance would be zero; if there are difference, the distance would always be a positive number. Negative distance? :confused: Couldn't comprehend.
—
All the best,
Shuanghe

Complete thread:

UA Flag
Activity
 Admin contact
23,428 posts in 4,929 threads, 1,690 registered users;
88 visitors (0 registered, 88 guests [including 7 identified bots]).
Forum time: 09:24 CEST (Europe/Vienna)

Anyone who conducts an argument by appealing to authority
is not using his intelligence;
he is just using his memory.    Leonardo da Vinci

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