On Thu, 26 Mar 2009 13:46:29 -0400, Jeffrey West
<[log in to unmask]> wrote:
>i am trying to set up contrast in the contrast manager. i have a 36 column
design matrix based on every 6 columns equal one of my conditions. ie: 1a 1b
1c 1d 1e 1f 2a 2b 2c 2d 2e 2f 3a... 6f. so if i want to combine all the a's, i
would enter into the contrast manager:
>1 0 0 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0
>it does not allow me to do this and i get a red line repeating these values
and it does not accept the contrast. i am not understanding since when i had
the same design matrix with 12 columns instead of 36 (1a - 2f only)
everything worked fine.
Some contrasts are not permitted because of a linear relationship among the
beta weights. In your case, I wouldn't think that would apply because of your
point that it worked for 12 columns.
If I recall correctly, you can investigate this possibility by using the SPM
facility for reviewing the design matrix. Underneath the design
matrix "schematic" is probably some line about the betas being uniquely
specified. If there's any gray in that line, this might be the problem.
>any help on this would be greatly appreciated.