<div>Dear All,</div>
<div> </div>
<div>I am running ICA on a dataset with two conditions focus on n400. ICA perform quiet well, but sometimes the N400 pattern appeared before ICA dataset was disappeared after ICA in single trial level. I am wondering whether this is because the common ICA weight matrix was used to fit the two conditions and mixed the residual error across the two conditions. Hence, I would like to use the common ICA weight matrix to run ICA on the two conditions separately, this may help me to estimate the activation of ICA component for each condition. However, I'm not sure how to do it in EEGLAB, and didn't find out the exact steps tutorial. Would somebody mind to help me? Or does somebody have other comments? Thank you very much!</div>
<div> </div>
<div>Best wishes,</div>
<div>Zude.</div>