> Could anyone point me in the right direction of a tutorial/any guidance on
> how to administer the parametric modulation analysis function in SPM2?
> I am uncertain what to enter at each of the prompts
The "ParStats" or "Parametric Model" in the example dataset should help:
http://www.fil.ion.ucl.ac.uk/~wpenny/datasets/face-rep/SPM2.html
(towards the end of the README.txt, it explains what buttons to press).
> I would also be interested to know people's opinions on the viability of
> using this analysis to analyse data from an N-back working memory paradigm
> - Can this really be considered a parametric variable?
I don't see why not.
Note that if N is small, you might find it easier to model each level of N as
a separate condition, rather than using a parametric modulation. You can
then test for any linear, quadratic, etc patterns in the contrasts (eg [-3 -1 1 3]
for a linear effect, if N=4). You will then find it easier to then plot the linear fit
together with parameter estimate for each condition (the latter is less "transparent"
when plotting using a parametric model instead).
The only real advantage of a parametric model would be if you don't want to
model higher-order effects. For example, you may know that only linear
effects of your parametric factor exist, in which case you could model your
4 levels with only 2 degrees of freedom (the mean and linear term). This would
be more powerful statistically. However, you would have to hope that your assumption
of no quadratic or cubic effects is true (otherwise the stats could be biased).
Note that if you did use a parametric modulation, and N=4, and you chose to
model linear, quadratic and cubic effects (ie 4 columns in total), your model
would have exactly the same explanatory power as if you had modelled each
condition separately (the F-tests across the 4 columns would be equivalent,
and you could always remap any contrast in one model to an equivalent
contrast in the other), ie the models would simply be "rotations" of each
other, and therefore "span the same space".
HTH
(ANTLA4ME)
Rik
----------------------------------------
Dr Richard Henson
MRC Cognition & Brain Sciences Unit
15 Chaucer Road
Cambridge
CB2 2EF, UK
Tel: +44 (0)1223 355 294 x522
Fax: +44 (0)1223 359 062
http://www.mrc-cbu.cam.ac.uk/~rik.henson
----------------------------------------
|