Hello!
I am still pretty new to SPM so hopefully my questions make sense.
I am
interested in running a 3 X 2 mixed-factor ANOVA (2nd level analysis) for my
fMRI data in SPM12. I know that using a flexible factorial design should allow
me to examine the interaction effect as well as the main effect for my
within-subjects factor, but I'm confused about how to properly enter the
contrast after I've run the estimation.
I have 45
subjects total, 3 groups (between-subjects factor; n = 15 per group) and 2
task conditions (within-subject/ repeated measure).
I've
specified
Factor 1 = subject
Factor 2 = group
Factor 3= task
Entered my matrix with 4 columns no problem with
Column 1 = 1s
Column 2= subject
Column 3 = group
Column 4 = task
So my factor
matrix =
1 1 1 1
1 1 1 2
...
1 15 1 1
1 15 1 2
1 16 2 1
1 16 2 2
...
1 30 2 1
1 30 2 2
1 31 3 1
1 31 3 2
...
1 45 3 1
1 45 3 2
I specified main effects for factors 1, 2, 3 and my interaction [2, 3].
So the column order I have is:
G1 G2 G3 T1 T2 G1T1 G1T2 G2T1 G2T2 G3T1 G3T2 and then 45 columns representing the 45 subs.
I read in a previous correspondence for a similar design that the contrasts to examine the interaction would be [0 0 0 0 0 1 -1 -1 1 0 0;-right padded with 45 zeros] and [0 0 0 0 0 0 0 1 -1 -1 1; right padded with 45 zeros].
I’ve run the interaction contrasts and got results just fine. To probe the interaction further, I was thinking I could run follow-up analyses, conducting paired t-tests for the effect of task within each of my three groups… is this a valid way to further explore my interaction effect?
I’m also unsure how to enter a valid contrast to look at the main effect of task (my within-subjects factor) in this design. I’m just wondering how to look at the main effect of task using my flexible factorial design as specified, if I’ve done something wrong, or if it would be best just to run a separate analysis using a paired t-test with all 45 subjects to just compare task 1 to task 2.
Any advice
would be greatly appreciated!
Thank you!
Theresa