Mean means [🇷 for BE/BA]

posted by ElMaestro  – Denmark, 2015-04-19 11:14 (4076 d 12:04 ago) – Posting: # 14706
Views: 32,926

Hi Helmut,

                                  R        T

❝ Method A (EMA, subjects fixed)   7.328141 7.391880

❝ Method B (EMA, subjects random)  7.328141 7.391880

❝ Method C (FDA = mixed efects)    7.328141 7.391880

❝ Marginal means (Gnumeric)        7.328141 7.391880

❝ So why do we get…

bear                             7.356274 7.420013

❝ … which are ~3.8% higher for both treatments? No, I’m not satisfied if only the PE agrees – I expect to get a nasty surprise once we start to deal with un­ba­lanced data.


You're right, it does look surprising.
Can you ask Phoenix what the log likelihood was when it had identified the minimum and compare with Yung-jin's logLik above?

And what about those DF's?

Pass or fail!
ElMaestro
Thread locked

Complete thread:

UA Flag
Activity
 Admin contact
23,654 posts in 4,992 threads, 1,571 registered users;
135 visitors (0 registered, 135 guests [including 14 identified bots]).
Forum time: 23:19 CEST (Europe/Vienna)

Always listen to experts.
They’ll tell you what can’t be done and why.
Then do it.    Robert A. Heinlein

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