In-vitro equivalence [General Sta­tis­tics]

posted by ElMaestro  – Denmark, 2012-05-22 16:08 (5137 d 00:01 ago) – Posting: # 8600
Views: 6,012

Hi Alex,

❝ Wouldn't that ignore the variability of the reference batch? It would be identical to an one-sample t-test to the reference PE for each batch. What if the focus is more on testing for equivalence than on testing for difference?


Absolutely correct re. the ignorance of ref. variability. However I am not sure in the context of this question that within-batch variability is the concern, given Martin's question "assuming normal distribution and independent replicates". In fact, when we look at between-batches (which I think is the likely aim for this QbD exercise) then the replicates within batches can certainly not be said to be independent or uncorrelated. In such situations, I think the theoretical way out is to use a mixed model. But it all depends on the purpose.
But as always I could be wrong, so...
Martin, do you wish to characterise the process performace within-batch or between-batch?

Pass or fail!
ElMaestro

Complete thread:

UA Flag
Activity
 Admin contact
23,654 posts in 4,992 threads, 1,570 registered users;
118 visitors (0 registered, 118 guests [including 11 identified bots]).
Forum time: 16:09 CEST (Europe/Vienna)

The idea is to try and give all the information to help others
to judge the value of your contribution;
not just the information that leads to judgment
in one particular direction or another.    Richard Feynman

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