Print

---

Hi

If there is no group x age interaction, I will use this model instead:

Group       EV1     EV2     EV3     EV4    EV5
1                1        0          0         0       20
2                0        1          0         0       30
3                0        0          1         0       40
4                0        0          0         1       50

Contrasts:   EV1     EV2     EV3     EV4    EV5
group by sex interaction
                  1         -1        -1        1         0
                 -1         1         1         -1        0
group effect
                   1        -1        1          -1        0
                  -1        1         -1         1         0
sex effect
                  1         1         -1          -1       0
                  -1       -1         1            1       0
age effect 
                    0        0          0          0        1
                    0       0           0          0        -1

So is it necessary to select the F test?