Bioequivalence and Bioavailability Forum • MSE tables

MSE tables [Software]

posted by ElMaestro – Denmark, 2014-02-24 19:17 (4496 d 12:38 ago) – Posting: # 12484
Views: 10,152

Hi Angus,

❝ ELmaestro: Thank you for your comments and encouragement. It seems that we have multiple linear regression used in conjunction with matrix alegebra....yes?

❝

❝ Would that be the case?

Indeed.

In matrix terms we use y=Xb+e

y is the vector of log-transformed metrics (logged observations).
X is the design matrix.
b is the effect vector (a vector of all the constants mentioned above with the redundant ones removed)
e is a vector of error.

Accordingly, max. likelihood is when e^te is minimised which gives us a useful b-vector.

—
Pass or fail!
ElMaestro

Complete thread:

Phoenix WinNonlin ANOVA AngusMcLean 2014-02-24 00:07
- Intercept Helmut 2014-02-24 00:51
  - Intercept AngusMcLean 2014-02-24 15:08
    - MSE tables Helmut 2014-02-24 15:58
      - MSE tables AngusMcLean 2014-02-24 16:56
        
        MSE tables ElMaestro 2014-02-24 17:26
        
        MSE tables AngusMcLean 2014-02-24 18:02
        
        MSE tablesElMaestro 2014-02-24 18:17
        
        MSE tables AngusMcLean 2014-02-24 19:56
        
        MSE tables ElMaestro 2014-02-24 23:58
        
        MSE tables AngusMcLean 2014-02-25 01:25