Hi,
You're right - this is possible with a combination of the ideas in the
FEAT manual examples. As well as having EVs for the main facto effects
and interactions (for interactions you typically multiply the
(hopefully already zero-mean!) relevant EVs together from the main
effects/factors). Also, don't forget to model the subject means (as
per the examples) so that you end up with the correct degrees of
freedom.
Cheers, Steve.
On 26 Feb 2009, at 18:09, Amy Krain wrote:
> I have data from 23 subjects on a task that has 6 different stimuli
> which represent a
> combination of two factors (probability and expected value). There
> are three levels of the
> probability factor and 2 levels of the expected value factor,
> resulting in the 6 stimuli. I am
> interested in conducting an ANOVA that would give me neural activity
> related to probability
> alone, expected value alone, and the interaction of the two. I am
> having trouble figuring
> how to do this- the examples in the FSL manual seem straightforward
> except I am unsure
> what the inputs are (copes from first level, I presume?) and they
> only use factors with 2
> levels so i am not sure how to make a similar model with 3 levels.
> Based on the FSL manual and previous posts, I thought I would need
> to create a design file
> with 23x6 inputs (copes for each stimulus type from individual
> feats). Then, each EV was a
> level of the first and second factors (resulting in 5 EVs). Once I
> did this, I realized that I
> didn't know how to create the interaction term. I have attached the
> design file that I
> started for 8 of the subjects.
> Any help with this would be greatly appreciated!
>
> Amy Roy, Ph.D.
> NYU Child Study Center
>
---------------------------------------------------------------------------
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
---------------------------------------------------------------------------
|