Hi - once you come to the randomise stage at the end then you can look up (on the FSL wiki and the email list archives) general queries about modelling with randomise - this isn't specific to dual-regression.

Cheers.

On 11 Feb 2013, at 08:09, Matthew Hyett wrote:

Dear FSL Users,

I am trying to find documentation on appropriate design.mat and design.con structure for a dual regression.  I have 3 groups (N=48) with resting state data that has been processed in Melodic.  At this stage, I am not looking at any covariates or group comparisons and simply wanted to inspect the individual spatial + temporal profiles.  As suggested in a previous post, dual regression may be a good option for doing this and I have looked at possible ways to construct the files with little success.

The randomise function appears to allow the user to specify a "-1" when treating the data as one group, which is, in short, what I was hoping to do.  Is this a valid shortcut in dual_regression?  Or will I need to specify the design.mat/.con files arbitrarily?

Cheers,
Matt

---------------------------------------------------------------------------
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
---------------------------------------------------------------------------