Hi all,

I haven't seen any replies on this yet, so I thought I would try posting again:

I am trying to analyze the effect of 2 covariates (both continuous variables) on 4 different emotion conditions in spm8.  I have tried setting this up in the 2nd level analysis, as full factorial, using one factor with 4 levels, then specifying my 2 covariates of interest.  I had to repeat each covariate 4 times to match each emotion condition within the same subject.

For each covariate, there is the option to create an interaction term with the factor (emotion).  In design1, I selected no for this option.  In design2, I selected yes which seems more appropriate.

Is this the correct way to essentially set up an Ancova in spm, or does anyone have a better suggestion?  Also, how would you set up the contrast manager to look at the effect of each covariate on each condition?


Ryan Herringa, MD, PhD
University of Pittsburgh