Reference and averaging

PB · Post by PB » Wed Jun 21, 2006 5:09 pm

Hi,

I just realized that when I average my EEG data (in order to get an ERP) I am using the unreferenced (i.e. CMS referenced) signal. First AFTER the averaging I choose an other channel as reference.

As far as my (incomplete) understanding goes, I have to do the re-referencing BEFORE making the average if I want to subtract the common mode correctly. Is this right?

Best,
Pär

Post by **Coen** » Fri Jun 23, 2006 12:53 pm

So long as you do not perform any artifact rejection, it does not matter whether the referring is done before or after averaging. However, performing artifact rejection on unreferenced (raw) signals will lead to wrong results. So, in practice you should always reference (and filter) the data before starting the averaging routines.

Best regards, Coen (BioSemi)