On Fri, May 14, 2010 at 2:31 PM, Dominik Bach <[log in to unmask]> wrote:
> I have a couple of minor questions about source reconstruction.
> First, Gareth recommended for the beamformers to use un-averaged data. But I
> noticed that at least for the imaging approach, you can also choose averaged
> data. Do you have a suggestion what is more appropriate/robust/sensitive?
If you are interested in evoked responses then beamforming won't work
so well and you should use SPM imaging solution. If you are interested
in induced responses you can use either and see what gives you better
> Then, the peristimulus time window is set in D.inv.contrast.woi - trials are
> often defined with respect to a trigger and can, e.g. be [-300 600], so is
> this with respect to the trial onset at -300 ms, or with respect to the
> trigger at 0 ms?
Times are always defined with respect to time 0 of the trial (trigger).