Hi group, I have some questions about the interpretation of my results based on the contrasts that i have used in my analysis. It is, i am sure a very basic question, and i would like to check if my interpretation of the results is correct/valid.
I have 2 groups, and 1 variable, and i am setting up a dods design, hence i have used 4 regressors. I query with a contrast vector of [0 0 1 -1], intending to discover regions where the measure x variable interaction between the groups is significant. Am i correct in the interpretation of the design and contrasts so far? In case yes, does the sign of the value in sig.mgh file indicate the direction of the interaction (i.e +ve values for positive correlation, and vice-versa) ? I realize that i am using "interaction" and "correlation" interchangeably: is this valid? How do i proceed from inferences about interactions to statements about correlation?
thanks, sid.