Dear SPMers,
I've just started trying to use the flexible factorial design spec. in SPM8. I have successfully built my model (matrix attached) - my question is how to build a contrast to look for a 3-way interaction. The SPM8 manual suggests specifying contrast for each two-way interaction then creating a further contrast to look for the difference between the two - is this correct, and if so how would this contrast be inputted.
Basically, I have 31 subjects;
Group 1; 13
Group 2; 18
Factor of Time with 2 levels each with 3 conditions:
1st level: Preoperative
Conditon 1
Condition 2
Condition 3
2nd level: Postoperative
Conditon 1
Condition 2
Condition 3
As in the matrix, you can see I modelled the main effect for subject, group, time and condition as well as the interactions for 'time x group' and 'time x condition'
I want to see if there is a 'group x time x condition' interaction.
Any help on contrasts would be much appreciated.
Many thanks
Jason
|