[Eeglablist] Discrepancy between STUDY results and manual averaging

[James Jones-Rounds]

Hello all,

Has anyone else noticed a discrepancy between ERPs generated with the STUDY
function (using EEGLAB v.9.x) and ERPs generated by loading all your
datasets and using "Plot-->Sum/compare ERPs" ?

Using Plot>Sum/compare ERPs we found an expected P300, increasing in
amplitude as we moved from frontal to parietal midline electrodes. Using a
STUDY design, however, we found at these same electrodes a strange pattern
of inverted polarity at adjacent electrodes (i.e. clear P300 at FPZ,
inverted P300, aka an "N300", at FZ,) back and forth down the midline.

Some notable points: a) we generated a STUDY design using participant data
files that were separated by condition (i.e. subj1_cond1, subj1_cond2,
subj2_cond1, subj2_cond2, etc.); b) the data were referenced to the average
of all electrodes (except the mastoid, which served as the original
reference). They were collected on a 64 channel NeuroScan unit with the
10-20 electrode positioning.

Thank you for any comments and suggestions you might have!

James

-- 
Laboratory Manager
Human Development EEG & Psychophysiology (HEP) Laboratory,
Department of Human Development,
--------------------------------------------
Technician and Laboratory Safety Officer
Computational Physiology Laboratory,
Department of Neurobiology & Behavior,
--------------------------------------------
Cornell University | Ithaca, NY
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20121019/ce8b1ba9/attachment.html>