Hi Erie,
What format did you use for specifying your "lower-level FEAT
directories" for your Lev3 analysis?

If you use something like:
path_to_S1.gfeat/cope1.feat
path_to_S2.gfeat/cope1.feat
path_to_S3.gfeat/cope1.feat
path_to_S4.gfeat/cope1.feat
...

it should work just fine.

cheers,
-MH

On Wed, 2011-05-25 at 12:50 -0700, Erie Boorman wrote:
> Hi Eugene, 
> 
> 
> Yep, you've got it right. I have thirty separate first level copes for
> each of four sessions and have ten subjects.  Following the
> instructions on the FSL webpage to the best of my ability, I first
> tried the following:
> 
> 1) Ran a lower-level FEAT for each run for each subject.
> 
> 2) Constructed a higher-level FEAT where the inputs were lower-level FEATs for each run for each subject (i.e. 40 FEATs corresponding to 40 runs).  In the stats tab, I selected the Fixed Effects option and set up the design with 40 EVs, where each EV picks out the 4 sessions that correspond to a particular subject. I also included 10 contrasts to represent the 10 subject means.
> 
> Rather than output something like subject_N.gfeat/cope1.feat, this second level analysis output 30 cope.feat directories, each with 10 cope.nii.gz files in its stats directory. The file structure was as follows: Group_Ana/cope[n].feat/stats/cope[j], where n = 1:30 lower level contrasts and j = 1:10 subjects.  This is different from what the instructions implied would be output.
> 
> Next, I followed some other FSL users' advice (and my intuition).  After completing step 1 above I:
> 1) Constructed a higher-level FEAT, using the fixed effects option, across the four runs per subject for each subject.  
> 2) Tried running a third level mixed effects analysis by selecting each person's feat directory corresponding to the fixed effects analysis across runs.
> When I set up the analysis, I receive the following warning message after selecting the Feat directory corresponding to the fixed effects analysis.  I receive ten of these messages (which corresponds to the number of subjects):  
> Warning: the first selected Feat directory contains no stats/cope images. 
> This is most likely because there are thirty cope.feat directories within the (.gfeat directory resulting from the fixed effects analysis), each with a stats directory and a single cope1.nii.gz within the stats directory.  It seems to fail because it cannot locate the stats directory within the .gfeat directories.  
> Cheers,
> -Erie
> 
> 
> On May 24, 2011, at 3:11 PM, Eugene Duff wrote:
> 
> > Hi Erie -
> > 
> > 
> > What are the errors (if any) that you're getting when you try the
> > approach described on the website?  Could you see anything in the
> > log files?
> > 
> > 
> > Have I got it right: you have thirty separate first level copes for
> > each session and you'd ideally want to follow the multi-session (4)
> > multi-subject (10) example for each?
> > 
> > 
> > Cheers,
> > 
> > 
> > Eugene
> > 
> > 
> > On 24 May 2011 21:56, Erie Boorman <[log in to unmask]> wrote:
> >         Hi,
> >         
> >         Thanks, Chris.  I followed your instructions and set up the
> >         third-level analysis to be run for each cope separately.  I
> >         ran this across copes using a bash script.
> >         
> >         While this worked fine it is hard for me to believe that
> >         there isn't a more automated way to do this?  For some
> >         reason, simply running a third level mixed effects analysis
> >         with the FE results by selecting each person's feat
> >         directory so that the analysis is done on all contrasts
> >         simultaneously does not work for me either.  This would be a
> >         lot more efficient/user-friendly.  In case it's of
> >         relevance, I'm using FSL 4.1 (2008).
> >         
> >         In any case, the instructions that are currently specified
> >         on the FSL webpage
> >         (http://www.fmrib.ox.ac.uk/fsl/feat5/detail.html#MultiSessionMultiSubject) listed under the section entitled Multi-Session & Multi-Subject (Repeated Measures - Three Level Analysis) appear to be incorrect.  Given how common it is to conduct an analysis using multiple runs per subject, it would be very helpful if the webpage could be updated.
> >         
> >         Cheers,
> >         -Erie
> >         
> > 
> > 
> 
>