[Eeglablist] epoch for ICA

[Tarik S Bel-Bahar]

Greetings Jinghao, 2 quick points below that might help, all the best!

1. When you are preparing your data for ICA, it can be epoched or not
epoched, it will not influence the general ICA results. The important thing
is that you feed ICA
enough data that is cleansed of big artifacts [please see eeglab wiki for
more information on what is "enough data for ICA" and what is "clean EEG
for ICA"].
Please also search on the eeglab list archives for previous discussions on
the topic of epoching for ICA.

2. the length of your epochs (or trials) is of course determined by your
experimental paradigm and the events of interest during the recording (such
as onset of a face on the screen that the participant is watching). It is
important for ICA that you only
include those EEG recording periods during the experiment that are of
interest to you,
and that are cognitively-relevant (the person is doing the task, not
sleeping or waiting).
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20130108/7ec44472/attachment.html>