If I had three groups (two subjects in each), and I wanted to run
an ANOVA, would this be appropriate?
EV1 EV2 EV3
1 0 0
1 0 0
1 1 0
1 1 0
1 0 1
1 0 1
My understanding is the F-test can be done like this?
C1 0 1 0 F1 x
C2 0 0 1 F1 x
The post-hoc contrasts could be done like this? If the
F-test came up significant?
C1 1 -1 0
C2 -1 1 0
C3 0 1 -1
C4 0 -1 1
C5 1 0 -1
C6 -1 0 1
Could this all be incorporated in to the same
model easily? For instance just make C1-C6,
C3-C8 and run it all together? Thanks again
for any feedback.
Chris Bell
|