Hi Nico,

Yes, what you described is the way to get what some people call 'pure
induced'. I think this is a rather artificial construct not
corresponding to any distinct physiological process because in reality
you get oscillatory activity with different degree of phase locking to
an event so it ends up in both evoked and induced components. So most
people just analyse the average TF and do not make this distinction.



On 30 May 2011, at 10:31, Nico Bunzeck <[log in to unmask]> wrote:

> Hi Vladimir,
> I am about to analyse MEG data regarding induced oscillatory power (at the
> sensor level; SPM8). I am not quite sure how to get there and would
> appreciate your help. If I understand the manual correct spm_eeg_tf followed
> by averaging (spm_eeg_average_TF) results in the power spectrum averaged
> over trials and this includes both 'evoked' and 'induced' power.
> Alternatively, averaging followed by spm_eeg_tf results in 'evoked' power
> only. Is that correct? Does that mean the difference between both would
> result in purely 'induced' power or is there another way?
> Thanks and best wishes,
> Nico