Dear Paul,
With regard to your learning experiment...
> I have conducted a learning experiment in a blocked design using fMRI.
> Learning proceeds across 6 blocks alternating with baseline. RT data were
> collected.
I would suggest that 6 scans are too few to allow you to model anything other
than a linear trend. It may be that you dont need to worry about modeling
exponential changes, since linearly increasing trends can account for most if
not all of the variance present in any non-linearly increasing trends in the
data. Also, if your RT data change in a roughly linear manner during learning,
you probably wont get significantly more out of the analysis than you would if
you just used a linearly changing model. I would just stick with a linear
model since you probably dont have sufficient resolution in the data to be
able to take advantage of the 'fine-tuning' methods that you suggest.
Best wishes,
Narender Ramnani.
University Laboratory of Physiology,
University of Oxford,
Parks Road,
Oxford OX1 3PT,
United Kingdom.
Oxford University Functional Magnetic Resonance Imaging Centre,
John Radcliffe Hospital,
Headington,
Oxford OX3 9DU,
United Kingdom.
Tel. +44 1865 222729 (Administrator)
+44 1865 222704 (Direct)
Fax. +44 1865 222717
email: [log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|