Conny,
I can;t think of any principled reason why you can't. As long as it
makes sense to test the null hypothesis across the two different
designs. So, can the response size in one design be sensibly compared
to those in the other? For example, comparisons between single event
and epoch-based designs are hard to interpret due to nonlinearities.
As well as the copes you'll need to scale the varcopes from the lower
level if you intend on using FLAME (and you ought to with different
designs at the lower level).
Cheers, Mark.
Dr Mark Woolrich.
Oxford University Centre for Functional MRI of the Brain (FMRIB),
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK.
Tel: (+44)1865-222782 Homepage: http://www.fmrib.ox.ac.uk/~woolrich
On 3 Jul 2006, at 11:49, Conny SCHMIDT wrote:
> Dear FSL users,
>
> is it possible to compare activation analysed with different
> designs? In
> detail we used a design convolved with the HRF and one, more epoch
> based,
> without convolution with the HRF. Since the peak-peak height of the
> EVs
> can be retrieved from the design.con files, I thought it might be
> sufficient to multiply the individual copes by these values and
> then run a
> second level, paired ttest analysis. Can we do so?
>
> Best regards and many thanks,
> Conny
>
> --------------------------------------------------
>
> Conny SCHMIDT, PhD
> Singapore Bioimaging Consortium (SBIC)
> BioMedical Sciences Institutes
> 11 Biopolis Way, #02-02 Helios
> Singapore 138667
> Fax: ++65 6478 9957
>
> --------------------------------------------------
|