Compilation [Study As­sess­ment]

posted by Helmut Homepage – Vienna, Austria, 2016-05-24 14:27  – Posting: # 16359
Views: 15,563

Dear all,

in R / lmerTest() one can get maximum likelihood estimation by setting the argument REML=FALSE in the model statement (default is REML=TRUE). Satterthwaite’s DF are 10.858 for the intercept and 6.884 for the slope. In lme() use method="ML" (default is method="REML"). Below a compilation (in analogy to BE rounded to four decimal figures and grouped by “similarity”):

Software        Method      Estimate      90% CI   
═══════════════════════════════════════════════════
SAS (Smith)   ML/Satter.   B0 1.94    1.54   2.35  
                           B1 0.7615  0.679  0.844 
SAS           ML/Satter.   B0 1.9425  1.5391 2.3458
                           B1 0.7615  0.6789 0.8441
R lmerTest    ML/Satter.   B0 1.9425  1.5391 2.3458
                           B1 0.7615  0.6789 0.8441
───────────────────────────────────────────────────
R lme         ML           B0 1.9425  1.5175 2.3675
                           B1 0.7615  0.6738 0.8491
───────────────────────────────────────────────────
SAS           ML/contain.  B0 1.9414  1.5175 2.3675
                           B1 0.7615  0.6738 0.8491
───────────────────────────────────────────────────
SAS           REML/Satter. B0 1.9414  1.4968 2.3860
                           B1 0.7617  0.6696 0.8539
PHX/WNL       REML/Satter. B0 1.9414  1.4968 2.3860
                           B1 0.7617  0.6696 0.8539
R lmerTest    REML/Satter. B0 1.9414  1.4968 2.3860
                           B1 0.7617  0.6696 0.8539
───────────────────────────────────────────────────
NCSS          method?      B0 1.9414  1.4849 2.3978
                           B1 0.7617  0.6659 0.8576
───────────────────────────────────────────────────
SAS           ML/contain.  B0 1.9414  1.4807 2.4020
                           B1 0.7617  0.6665 0.8570
R lme         REML         B0 1.9414  1.4807 2.4020
                           B1 0.7617  0.6665 0.8570
───────────────────────────────────────────────────
PHX/WNL       REML/resid.  B0 1.9414  1.4515 2.4313
                           B1 0.7617  0.6665 0.8570


My preference is REML/Satterthwaite because one could reproduce results in three different software packages. I would avoid NCSS; no idea how the calculation is done.

Cheers,
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. ☼
Science Quotes

Complete thread:

Activity
 Mix view
Bioequivalence and Bioavailability Forum |  Admin contact
19,537 posts in 4,144 threads, 1,338 registered users;
online 5 (0 registered, 5 guests [including 5 identified bots]).
Forum time (Europe/Vienna): 17:49 CEST

When puzzled, it never hurts to read the primary documents –
a rather simple and self-evident principle that has, nonetheless,
completely disappeared from large sectors
of the American experience.    Stephen Jay Gould

The BIOEQUIVALENCE / BIOAVAILABILITY FORUM is hosted by
BEBAC Ing. Helmut Schütz
HTML5