Dear Bartosz,
>We are planning a working memory study with visual presentation of items
>consisting of either:
>
> rest condition(<24s) - control condition with low wm- load (24s) -
>activation condition(24s) - rest condition ....
>or: rest - control - rest - activation - rest - control....
>
>Is it possible to use the first (reduced) design keeping the possibility to
>make the "same" contrasts in SPM (i.e. control vs. rest, activation vs. rest,
>activation vs. control...)? I think no.
The first design is fine: you are not restricted to interposing a rest
condition between every other condition. Placing a longer interval between
repetitions of the same (rest) condition makes you slightly more vulnerable
to low frequency drifts, but for practical purposes with the length (24s) of
your epochs this is not a problem here.
>Does it make sense to use the rest-condition as instruction-condition in order
>to avoid popping out the instruction with beginning of the blocks?
I am not sure what you mean by 'instruction-condition' here. If this means
subjects are simply briefly cued at the end of the rest period as to which
active condition to expect, this would be fine. But if it means some sort of
continuous instruction throughout the 'rest' period, then this of course
means 'rest' is no longer 'rest' but 'instruction'.
>Modelling the rest condition in SPM 99b: How to avoid the mesurement of
>activation-changes from activation to rest and not the rest condition itself (
>because of this a rest-length of under 15-20s might be too short!?)
I am not sure I have understood what you mean here - perhaps you could
clarify? If you use a boxcar model in SPM99b, you will be estimating the
mean activation at each voxel for each different condition. Linear contrasts
between each condition will assess changes in activation of experimental
interest. Your rest condition is a little short for a blocked design, but
not dangerously so (>16s fine).
Best wishes,
Geraint
---------------------------------------------
Dr. Geraint Rees
Wellcome Advanced Fellow,
California Institute of Technology,
Pasadena,
CA 91125
voice (626) 396-2880
fax (626) 796-8876
web http://www.klab.caltech.edu/~geraint
---------------------------------------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|