Hi Anderson,
The following is how the design and contrast files look like
1. design.mat (first column = group 1; second column = group 2; third column= age; fourth column= gender)
================================
/NumWaves 4
/NumPoints 104
/PPheights 1 1 42 1
/Matrix
1 0 26 1
1 0 32 1
1 0 42 1
1 0 22 1
1 0 25 1
1 0 24 1
1 0 24 1
1 0 32 1
1 0 25 1
. . . .
. . . .
. . . .
================================
2. design.con (I am trying to find the group difference between group 1 and group 2 with age and gender as nuisance variables)
================================
/NumWaves 4
/NumContrasts 2
/PPheights 1 1 0 0
/Matrix
-1 1 0 0
1 -1 0 0
================================
3. design.fts
================================
/NumWaves 2
/NumContrasts 2
/PPheights 1 1
/Matrix
1 1
================================
In my hypothesis, i am trying to ask the question if all the three input modalities together can explain a significant group difference? I was wonderding if i can use simple MANOVA to answer this question. Or should i use MANOVA with contrasts (since i have 2 contrasts in design.con) in this context?
Thank you
Jay
|