Hi - because there is so much evidence now that applying Granger to
BOLD data is potentially dodgy, I'm afraid we don't have direct
experience of the likely interactions of these processing steps,
including slice timing correction (which will induce different
temporal smoothness to different slices), smoothing, and whitening.
There certainly will be some interactions, though.
Cheers, Steve.
On 29 Aug 2009, at 13:56, Leij, A.R. van der wrote:
> Hi,
>
> I'm preparing a dataset for a ROI-based granger causality analysis (in
> Matlab). I will preprocess the data with slicetime and motion
> correction
> and will try out no spatial smoothing and 6mm spatial smoothing.
>
> I was wondering the following: would it be allowed to turn on
> prewhitening prior to a granger causality analysis? If I would do
> so, I
> can imagine this will result in far more robust estimates of temporal
> (auto)correlation between successive dynamics because you eliminate
> the
> temporal autocorrelated noise as much as possible. However, since I'm
> not entirely aware of the underlying details of both methods (which
> are
> probably related), I'm a bit worried that this will be a statistically
> invalid/illegal approach, fi because of artificially lifting temporal
> relations.
>
> Thank you!
>
> Andries van der Leij
>
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
|