Hi,
In my experiment I have two groups of subjects (patients (n = 20) and controls (n =42)) performing a memory task with 2 conditions. I'm interested in main effects of group, conditions and interactions between group and conditions. What is the most appropriate design to answer these questions?
1) Full factorial design with 2 factors: group (2 levels), independence (yes), variance (unequal); conditions (2 levels), independence (no), variance (equal). To evaluate the differences I used the following F-contrasts:
Main effect of group: 1 1 -1 -1
Main effect of condition: 1 -1 1 -1
Interaction group x condition: 1 -1 -1 1
2) Flexible factorial design with 3 factors: 1) subject, independence (yes), variance (equal); 2) group, independence (yes), variance (unequal); 3) condition, independence (no), variance (equal). In main effects & interactions: main effects (1,2,3) and interaction (2,3). To evaluate the differences I used the following F-contrasts:
Main effect of group: 1 -1 0 0 0.5 0.5 -0.5 -0.5 ones(1,20)/20 -ones(1,42)/42
Main effect of condition: 0 0 1 -1 0.5 -0.5 0.5 -0.5 zeros(1,20+42)
Interaction group x condition: 0 0 0 0 1 -1 -1 1 zeros(1,20+42)
Is one of these two approaches better than other?
Thank you very much in advance,
Claire
|