Hi,

It's the same thing. When you choose the "conditions of interest F-contrast" to correct your eigenvector, it demeans the eigenvector by all effects NOT included in that contrast (i.e. the effects of no interest).

I know, the wording is rather confusing...

So, if you have 2 regressors of interest followed by regressors of non-interest, you should choose an F-contrast of eye(2) when you are asked to correct your data. This will demean your eigenvalues by all regressors of non-interest.

On Thu, Sep 3, 2015 at 4:21 AM, Frederik Van de Steen <[log in to unmask]> wrote:

Dear DCM experts,

In many papers and list archives, I have read that when extracting time series for DCM, you have to adjust for effects of interest (F-contrast). However, in a recent paper by Vossel et al (2015, Journal of neuroscience, with Prof. Friston as co-author), they corrected for effects of no-interest:

"The resulting time series were adjusted for effects of no
interest (e.g., rest periods, error trials) and physiological confounds so that
the analyses focused on BOLD responses reflecting the effects of valid and
invalid cueing and the modulation by precision (effects of the volatility re-
gressor were also excluded)."

So my question is: When do you correct for effects of interest and when do you correct for effects of no-interest (and why)?

Kind regards,

Frederik Van de Steen