Desultory thoughts [🇷 for BE/BA]

posted by ElMaestro  – Denmark, 2020-07-16 01:06 (1371 d 17:17 ago) – Posting: # 21705
Views: 16,533

Hi Hötzi,

❝ ❝ to me it is exactly the opposite way around:

❝ ❝ A model with more than one variance component is a mixed model.


❝ Not necessarily. It depends on what you believe [sic] is a random effect. Treatment, sequence, and period are fixed effects, right? IMHO, subject is random. When we think about interaction(s) we enter the gray zone, of course.


May I offer the completely opposite view? Search the forum for "bogus statement" :-D, check the SAS documentation for what the statement actually actually does to Proc GLM:
When ProcGLM is used for 222BE with or without the bogus statement it still means a normal linear model is fit (even when the theory in e.g. C&L calls it "random"). There is a single variance component in such fits. If you wish to verify it: There is a model matrix with columns for subjects; check the df's for such models, it would be (much) higher if subjects were fit as random. Conversely, if subject were random subject would not appear as a factor in the anova.

Pass or fail!
ElMaestro

Complete thread:

UA Flag
Activity
 Admin contact
22,985 posts in 4,823 threads, 1,655 registered users;
45 visitors (0 registered, 45 guests [including 8 identified bots]).
Forum time: 18:24 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