On Thu, 6 Dec 2007 14:56:05 +0000, Markus Burgmer <[log in to unmask]
MUENSTER.DE> wrote:
>dear spmers.
>
>i performed a simple block design with 3 conditions (a,b,c). the
conditions
>increase in their stimulation strength and each contains 5 scans. each
>condition was repeated 5 times in a pseudorandomized way, so i got 15
scans
>for each condition. at the beginning i included one baseline condition of
5
>scans.
>my question is:
>is it allowed to include the single baseline condition (bc)of 5 scans as
>the rest condition in my first-level model while it contains only a fifth
>of the scans of each stimulation condition?
>or do all conditions which were contrasted at first-level must contain
the
>same number of images?
I've only given quick thought to the details of your design, but a good
rule of thumb is that the number of scans per condition (where baseline is
also considered as a condition) should be roughly equal, if the conditions
are all equally important in subsequent contrasts.
On the other hand, you might not "care about" some of the conditions as
much as others.
Besides the simple number of scans per condition, there's also the issue
of the arrangement of the conditions. In event-related fMRI, where
the "blocks" are extremely short, this is very important. In block-design
fMRI, you still have to consider whether your design is confounded with
low-frequency drifts. (E.g. a design that looked like A A A A B B B B
could be argued to be highly confounded with drift.) This latter issue is
probably not relevant to your design if you've mixed things up enough.
>
>thanks in advance
>markus
>========================================================================
|