Hi - yes, that seems right I think - though note that strictly
speaking statistically, finding no significant group difference is not
the same as an explicit test that the null is true....
Cheers.
On 18 Mar 2009, at 12:59, Daniel CM wrote:
> Hi I just wanted to check that we are doing the correct thing here...
>
> We scanned an initial group of subjects on an fMRI task and we
> realised that
> there was a problem with the aquisition sequence. So, we scanned a
> new
> group of subjects again, a few months later. Of course, we wanted to
> see if
> we could still use the original subjects. Our strategy was as follows:
>
> Place all the lower level feats into a higher-level analysis as two
> groups, in
> two columns as outlined in the Feat manual. [# #]. Then check for
> differences between the groups [1 -1] and [-1 1] on each contrast.
> Since
> there were no major differences, but still thinking it is safer
> treat these groups
> as having different variances, we then wanted to know the COPE for
> both
> groups combined. So, we simply set the higher level contrast as [1 1].
>
> Does this sound about right?
>
> Cheers,
> Dan CM
>
---------------------------------------------------------------------------
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
---------------------------------------------------------------------------
|