Dear Joseph,
Yes, this is possible.
If your two covariates are c1 and c2 then create a third covariate
ix=(c1-mean(c1)).*(c2-mean(c2))
Add this new covariate into your model and estimate the new parameters.
A contrast of [0 0 1] will then look to see where BOLD increases when
both c1 and c2 are above their mean (or where it decreases when both c1
and c2 are below their mean).
A contrast of [0 0 -1] will look for the reverse pattern.
Best, Will.
Joseph Beeney wrote:
> Is it possible to look at areas of brain function on a task that were
> modulated by an interaction of two covariates in a way that would be
> interpretable (e.g., high x, high y modulated activity in these areas)?
>
> Joseph E. Beeney, M.S.
> Doctoral Candidate
> Department of Psychology
> Pennsylvania State University
> 545 Moore Building
> University Park, PA 16802
--
William D. Penny
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG
Tel: 020 7833 7475
FAX: 020 7813 1420
Email: [log in to unmask]
URL: http://www.fil.ion.ucl.ac.uk/~wpenny/
|