Hi Laura,
> Dear SPMers,
>
> I am using SPM2 to analyse some fMRI data, and I have got a doubt on
> the choice of a couple of parameters when doing the
>
> fMRI/specify data
> SPM asks me:
> "remove global effect", and the choice is 'none' or 'scale'.
> Which is the difference? Are there some case in which one choice is
> better that the other?
>
The current tendancy is to select 'none', but it is still controversial
.. please have a look at the archive of this year with "Global Scaling"
keywords....
> Also, when SPM asks for the highpass filter, and I choose 'specify'
> the default value is 128, but is there a way of choosing an optimum
> parameter for the experiment considered? Someone told me that if for
> example the experiment is a block design with eight rest and eight
> activations interleaved, then one can choose the high pass filter as
> corresponding to the length of four cycles of rest-activation. Is this
> reasonable or better than using the default value?
>
The goal is to remove low frequency variations (which are also often
related to global effect). According to your design you can select
different frequencies, one can use 2.5 x maximum SOA lengh ..
Best,
- cyril
|