Replicate Design ≠ Higher-Order Cross­over [Design Issues]

posted by Relaxation  – Germany, 2023-03-20 09:56 (830 d 18:32 ago) – Posting: # 23506
Views: 5,633

Hi everybody and sorry for posting this additional question late.
But I cannot get to a clear answer by myself.

Helmut, you stated

❝ What you must not use is the Latin Square ABCD|BCDA|CDAB|DABC. Only from a Williams’ design you can extract balanced pairwise comparisons


I somehow fail to see why the extracted "pairs" are not considered balanced.
They are not balanced for carry-over (which was always the most important reason for Williams), but aren't they for period and sequence?
Is this what you mean here?

I have no idea how to make it fancy, so I will try an ASCII-style
Sequence | Treatments | IBD for A vs B 
1        | ABCD       | AB..
2        | DABC       | .AB.
3        | CDAB       | ..AB
4        | BCDA       | B..A

So A is present once in each sequence and period and the same for B.

Best regards,
Relaxation.

Complete thread:

UA Flag
Activity
 Admin contact
23,424 posts in 4,927 threads, 1,685 registered users;
37 visitors (0 registered, 37 guests [including 9 identified bots]).
Forum time: 05:28 CEST (Europe/Vienna)

Complex, statistically improbable things are by their nature
more difficult to explain than
simple, statistically probable things.    Richard Dawkins

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