Dear Anderson
So there are 4 groups: male patient (EV1) male control (EV2) female patient (EV3) female control (EV4)
I then get the mean of ALL participants' ages and demean each participant with the resulting (EV5)
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
Now the design becomes this (with the consideration of group x age interaction):
Group EV1 EV2 EV3 EV4 EV5 EV6 EV7 EV8
1 1 0 0 0 20 0 0 0
2 0 1 0 0 0 30 0 0
3 0 0 1 0 0 0 40 0
4 0 0 0 1 0 0 0 50
Contrasts: EV1 EV2 EV3 EV4 EV5 EV6 EV7 EV8
group by sex interaction
1 -1 -1 1 0 0 0 0
-1 1 1 -1 0 0 0 0
group effect
1 -1 1 -1 0 0 0 0
-1 1 -1 1 0 0 0 0
sex effect
1 1 -1 -1 0 0 0 0
-1 -1 1 1 0 0 0 0
So my question is this:
Despite this change of design, I think the results of "group by sex interaction", "group effect" and "sex effect" should remain the same?
|