Hi Niels,
On Mon, 5 Sep 2005, N.M. van Strien wrote:
> Hi,
>
> I have a question about setting up my analysis for a block design task. I
> have two conditions A and B, and a fixation cross in my experiment. For
> technical reasons I can run only blocks A – fix – A … etc , B – fix – B …
> etc. The main comparison I want to make is the comparison between the
> activation of A and B.
>
> I have (for now) 1 pilot subject who was scanned under condition A and B.
> Condition A consists of 4 runs (A – fix A etc). Condition B also consists
> of 4 runs (B – fix B etc).
>
> Since I wanted to pool my data over condition, I started my analysis by
> doing first level feats for both the 4 A runs and B runs, giving 8 first
> level outputs. As EVs I defined:
> A
> Fix
> A > fix
I'm not sure what you mean by this - what you want is a single EV that is
(eg) 0 for the fix timepoints and 1 for the A timepoints and then a single
[1] contrast to pull out the A-Fix parameter estimate.
> And similar for B.
>
> In the second level analysis I grouped my data by performing 2 higher
> level analyses, taking the four A or B first level analyses as inputs. The
> contrast I chose consisted of one EV which I set to 1. I think this will
> give me the mean for each of the above contrasts (resulting in 2 gfeat
> dirs and 6 feat subdirs).
>
> As a final step I wanted to use the feat subdirs from the gfeat dirs from
> my second level analysis as inputs. Now I am bit puzzled, since I am not
> sure how to define this contrast properly.
I don't think I would do it like this - I would feed all 8 lower-level
analyses into a single second-level analysis, which would look like the
two-sample unpaired t-test example shown at
http://www.fmrib.ox.ac.uk/fsl/feat5/detail.html#UnpairedTwoGroupDifference
Note that for this within-subject analysis you may not be interested in
the cross-session variance, in which case running the second-level
analysis as fixed-effects instead of mixed-effects is valid and will give
you bigger blobs!
Cheers, Steve.
--
Stephen M. Smith DPhil
Associate Director, FMRIB and Analysis Research Coordinator
Oxford University Centre for Functional MRI of the Brain
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
|