Dominik,
I don't have any specific advice or suggestions, but I would think that the
appropriate analysis would depend on the minimum number of stimuli per
class, the nature of your stimuli, and what your hypothesis is. For
example, are you expecting differential responses to different stimuli, a
response to some but not all stimuli, or a correlation with behavioral
responses to one or more class of stimulus? It's hard to give suggestions
without knowing more details about the experiment and what you are trying
to show.
Doug
At 04:40 PM 11/16/2006 +0100, Dominik R. Bach wrote:
>Hi all
>
>I want to analyse my fMRI data according to behavioural measures, i.e.
>stimuli are to be classified by the subject's response. This leads to an
>extremely unbalanced design, where the number of stimuli per class differs
>between subjects by a factor of up to 3.
>
>How does this affect two-level-analysis? I read that in this case it would
>be more appropriate to use a maximum-likelihood-approach in a one-level
>analysis for all subjects. Is this possible within SPM?
>
>Thanks
>Dominik
Dept. of Communication Sciences & Disorders
Northwestern University
2240 Campus Drive
Frances Searle Building, Room 2-356
Evanston, IL 60208
phone 847-467-1549
fax 847-491-4975
email: [log in to unmask]
|