Dear Ariadna,
Either (a) don't model error for these participants or (b) create a
dummy condition with a single trial at the end of the session. If you go
for (b), e.g. because it preserves the dimensions of the design matrix
for all participants, make sure to always have a 0 weight in contrasts
for that condition and scale contrasts appropriately if necessary.
Best regards,
Guillaume.
On 14/03/2019 10:37, Ariadna Albajara Saenz wrote:
> Dear SPM experts,
>
> I have five regressors for my task, one of which models the errors. However, not all participants have errors . Do you know if in those cases I should enter only 4 regressors and eliminate the error regressor?
>
> Thank you in advance!
>
> Ariadna
>
--
Guillaume Flandin, PhD
Wellcome Centre for Human Neuroimaging
UCL Queen Square Institute of Neurology
London WC1N 3BG
|