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/ . =================================================================
