Hello,
I have run a 2x3-ANOVA (within subject) using the flexible factorial option with two main factors A, B (task, condition) and the corresponding interaction.
I followed the instructions by Jan Gläscher and Darren Gitelman in their excellent tutorial "Contrast weights in flexible factorial design with multiple groups of subjects".
Now, my factor B consists of three levels B1 B2 B3 as well, but in contrast to their example, I do not have a linear increase (like 1-back < 2-back < 3-back) or better, no increase at all.
To make sure, is it correct then to set up F-contrasts for main effect A:
1 -1 0 0 0 1/3 1/3 1/3 -1/3 -1/3 -1/3
instead of
1 -1 0 0 0 1/3 0 1/3 -1/3 0 -1/3
and for main effect B
0 0 1 0 -1 .5 0 -.5 .5 0 -.5
0 0 1 -1 0 .5 -.5 0 .5 -.5 0
instead of
0 0 1 0 -1 .5 0 -.5 .5 0 -.5
and for interaction
0 0 0 0 0 -1 0 1 1 0 -1
0 0 0 0 0 -1 1 0 1 -1 0
instead of
0 0 0 0 0 -1 0 1 1 0 -1
Second question, if I want to compare B1 with B2, then the corresponing vector is
0 0 1 -1 0 .5 -.5 0 .5 -.5 0
In case I wanted to compare A1B1 with A1B2 only inside the model for inspection, is there a possibility at all? How would such a vector look like?
Many thanks for help,
Helmut
|