 |
 |
Dear SPMers I have a question concerning DCM. I looked through all posts, but could’nt find an answer yet. I’m using SPM 8 with DCM 10. In my fMRI experiment, I have 3 Conditions A, B and C. For my DCM analysis I want to set up a design matrix with 4 columns: column 1 : All conditions (A, B and C) column 2 : condition A column 3 : condition B column 4 : condition C Because column 1 is a linear combination of condition A, B and C, the beta for this regressor is not uniquely specified, i.e. appears gray in the estimability. Column 1 seems to be important for the input in the DCM. But since the parameter estimates are not uniquely specified, I can’t calculate any F-Contrast. I did the DCM-tutorial in the SPM manual. In this example the first column is a linear combination of the condition too, but parameter estimates are unique. You can see a comparison of the design matrix in the DCM-tutorial and my design matrix. What am I missing? What do I have to change to do make f-contrasts possible? Best regards Anne