Bioequivalence and Bioavailability Forum • PowerTOST robust

PowerTOST robust [Power / Sample Size]

posted by d_labes – Berlin, Germany, 2012-01-17 12:53 (4851 d 07:28 ago) – Posting: # 7956
Views: 9,160

Dear Ben,

❝ First, I want to say thanks for the great package PowerTOST!

Thanx for the flowers.

❝ I saw that you added some new features, in particular the df2 degrees of freedom. The latter brings me to my second point. In your excerpt (or the manual) you wrote "The df2 are also more appropriate if the planning of sample size is done based on CV’s originating from real mixed model analysis (via Proc MIXED in SAS or lme() in R)". Why is that? Is there a reference where this is mentioned (haven't seen it explicitly in Senn's "Cross-over Trials in Clinical Research")?

Sorry. No reference available. It's my personal observation looking at results of the FDA code for replicate studies (which uses Proc MIXED as you know).

—
Regards,

Detlew

Complete thread:

Power for unbalanced cross-over d_labes 2011-01-14 15:21
- Great post! Helmut 2011-01-14 16:16
  - Santa is coming d_labes 2011-12-22 09:30
    - PowerTOST Ben 2012-01-02 17:39
      - BIBDs Helmut 2012-01-02 17:53
        
        Uuups d_labes 2012-01-17 11:49
        
        the \tests subdirectory in a R-package yjlee168 2013-06-11 23:03
        
        To demo() or not d_labes 2013-06-14 13:28
        
        demo() is not demo yjlee168 2013-06-15 22:04
        
        OT: issue of '\tests' subdir again yjlee168 2013-09-28 23:23
      - PowerTOST robustd_labes 2012-01-17 11:53