[Eeglablist] Measuring clusters using numerical integration/ area under the curve

[Elizabeth J Kirkham]

Dear all,

I am using the area under the curve methodology to measure my ERP
waveforms. To do this, I am using ERPLAB's Measurement tool function, with
the settings "Numerical Intergration/Area between two fixed latencies",
"Rectified area (negative values become positive)".

I want to examine the waveform over two clusters of electrodes (P5, P7 and
TP7; and P6, P8 and TP8).

There are two ways for me to examine these clusters, either (1) create the
cluster in ERPLAB (ERPLAB > ERP operations > ERP channel operations), and
then use the Measurement tool to find the Numerical Integration values for
the cluster, or (2) use the Measurement tool to find the Numerical
Integration values for each electrode separately, and then average them
together later (for example in Excel).

I have found that these measures produce different output values. The
values are not the same in absolute terms, or in relative terms. (For
example, the values for one participant in conditions 1 and 2 are 0.523 and
0.630 using the first method, and 1.698 and 1.641 using the second method.)

I don't know if these different values are because I have made a mistake,
or if differences in output are to be expected when using numerical
integration?

If it is the case that I have not made a mistake, is there a best practice
in terms of which way round to carry out the procedures? Should I average
the channels in ERPLAB, and then take the measurement, or take the
measurement for each channel separately, and then average the output
together?

Thanks in advance,
Elizabeth


-- 
Elizabeth Kirkham
PhD Researcher

Department of Psychology
The University of Sheffield
Floor E
Cathedral Court
1 Vicar Lane
Sheffield
S1 2LT
UK
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20161205/db7d59b1/attachment.html>