Hi .
Working on designing an analysis for a repeated measures ANOVA.
I know, I should probably not use ANOVa, but my colleagues want to see the result.
Anyways, I have 10 subjects that all underwent 5 conditions under each of 3 tasks, and I am
tryiing to set up the mean TASK, mean CONDITION, and interactions. Thus, I have 150 input copes
(10 subjects, 3 tasks, 5 conditions)
so far I have
EV1 through EV10 being the input levels of each subject (i.e., all 1's for a given subject regardless
of task and condition, and 0's everywhere else).
EV11 is the mean of the copes for TASK.
EV12 if the mean of the copes for condition.
My question is how to model the interaction? I can intuitively follow how to do it in a non-repeated
design with only 2 factors, bt having trouble with any more than that.
Do you need to model some sort of linear trend across tasks or conditions to do this?
-Brad
|