Imbalance or incomplete data, SASophylistic view [🇷 for BE/BA]

posted by d_labes  – Berlin, Germany, 2008-11-11 15:02 (6062 d 11:13 ago) – Posting: # 2655
Views: 33,604

Dear Yung-ying,

❝ still has ANOVA Table with PROC MIXED? ;-)


The heading you cited is from Proc GLM :-P.
Under Proc MIXED there are no SSq nor MSq.
This is because the REML estimation method does not rely on SSq's and their decomposition.

❝ What 90%CI method did you use?


The shown point estimates and 90% CI's are from the difference between Least square means.

Here is my code for Proc GLM / mixed (to play around with, it seems you have access to "The power to know"):
ODS output LSMeandiffCL=ratios; *<- this saves the 90% CIs;
Proc GLM data=BEBAC24long;
  class formul period subject sequence;
  model logCmax=formul period sequence subject(sequence);
  random subject(sequence) /test;
  lsmeans formul/cl pdiff alpha=0.1;
run; quit;
*log diff R-T is calc. because of lexicographic order, but we need T-R;
Data ratios;
  set ratios;
  point=Exp(-difference);
  lower=exp(-upperCL);
  upper=exp(-lowerCL);
run;


ODS output Diffs=ratiosMixed;*<- this saves the 90% CIs;
Proc mixed data=BEBAC24long;
  class formul period subject sequence;
  model logCmax=formul period sequence /ddfm=satterth;
  random subject(sequence) /subject=subject;
  lsmeans formul/cl diff alpha=0.1;
run;
—
Regards,

Detlew

Complete thread:

UA Flag
Activity
 Admin contact
23,424 posts in 4,927 threads, 1,673 registered users;
25 visitors (0 registered, 25 guests [including 16 identified bots]).
Forum time: 03:16 CEST (Europe/Vienna)

Truth and clarity are complementary.    Niels Bohr

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