If I've counted correctly, you have 15 conditions + 6 movement
parameters (21 total) per session. So can you check that the
contrasts you specify have the appropriate number of columns? I.e.
they should have 168 (8 sessions * 21 columns) + 8 (for the session
effects) = 176 columns each. If so, can you respond with the exact
contrast you've entered?
On Thu, May 6, 2010 at 2:57 AM, Yune Lee <[log in to unmask]> wrote:
> Hi Jonathan,
> Many thanks for the help.
> Enclosed is the design matrix.
> On Wed, May 5, 2010 at 8:43 PM, Jonathan Peelle <[log in to unmask]> wrote:
>> Hi Yune
>> > I'm having trouble creating con images and I suspect this may have
>> > something to do with the identical design matrix across runs.
>> > So, in my study, I have four runs of visual conditions followed by four
>> > runs of auditory conditions. (They were collected on separate days, but
>> > were combined for multi-modal analysis) The order of contrast values
>> > are
>> > exactly identical across runs (e.g., 1, 1, 0, 0 (run1), 1,1,0,0 (run2)
>> > ..-1-1,0,0 (run7), -1,-1, 0,0 (run8).
>> > Is there any way to resolve this 'invalid contrast' issue?
>> From what you describe I can't see any problem with your design matrix
>> (unless I'm missing something)—it's common to have the same design
>> across different sessions.
>> Can you send a picture of your design matrix? Any chance the problem
>> is something like not specifying the correct number of values (one per