Hi,

Suppose one has a 3 x 2 factorial ANOVA design with homogenous
variances, equal n's, and individual conditions labeled as follows:

A B C
D E F

Can the R2 for an overall model (either a reduced main effects model
or full interactive model) be determined from the R2's for the three
possible 2x2 submatrices (e.g., R2a, R2b, R2c)?

A B
D E

B C
E F

A C
D F

I'm proceeding by deriving the individual correlations, but am
thinking there may be some existing work on this that I may have
missed.

Thanks for any advice,

-wp
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to