Open issues [🇷 for BE/BA]

posted by ElMaestro  – Denmark, 2020-08-06 23:11 (1588 d 15:57 ago) – Posting: # 21827
Views: 20,661

Hi PharmCat,

I really like your idea. I don't have a good understanding of this as a whole but I know that when I work with a little these things they become easier for me to grasp, so I am giving it a try. :-)

Let us turn to Gurka 2006 equation 3 and 5 and remembering that the rank of our overall design matrix X is 6:
Gurka requires me to derive Xi, the design matrix for fixed effects of subject i.
They mention that "Xi is a (ni × p) known, constant design matrix for the i' th subject
with rank p"

At just three observations per subject we cannot have a rank of p=6 for Xi. The maximum rank is naturally 3. I don't see a way to handle Xi on a subject by subject basis, but perhaps you have some insight. Possibly I could do it by lumping subjects together so that the "smaller" X of sorts still has more rows thank columns and thus a rank of p=6.


Do you see my issue with Xi? Could you indicate your view on Xi here? Is lumping subjects together a way to go to ensure the right rank of the reduced design matrices? Many thanks in advance.

Pass or fail!
ElMaestro

Complete thread:

UA Flag
Activity
 Admin contact
23,336 posts in 4,902 threads, 1,669 registered users;
38 visitors (0 registered, 38 guests [including 9 identified bots]).
Forum time: 14:08 CET (Europe/Vienna)

Biostatistician. One who has neither the intellect for mathematics
nor the commitment for medicine but likes to dabble in both.    Stephen Senn

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