help with analysis - EEG data
Elisa
I need some help with my analysis as I am just now starting to play
around with this toolbox.
I have EEG data, with 60 channels. In my experiment I have 2 types of
events that I want to compare and therefore see whether there are any
differences in responses. Within the time series, there are 39 epochs
for each event, for each subject, running across all electrodes.
Therefore I have a time series with 78 epochs. I was considering
dividing the time series in two, one for event 1 and one for event 2,
and then compare among electrode pairs for each event type. This means
that I will look at electrode 1 vs electrode 2 for event type 1, then
for event type 2, and so on with all electrode pairs.
1) Do you think this could be correctly done by using the event-
related type of analysis? If not, should I then use the whole time
series approach to go about looking at the differences of correlation
between pairs of electrodes across different events?
2) If the event-related design can be used, what should I put in the
"scans per trial" input, considering that I'm using EEG data ?
3) Also, the vector of stimulus time-series is the times at which the
stimulus of interest occurs across the whole time series?
If 3 is correct, then I assume that I cannot average across all 39
epochs before running the analysis, in order to end up with an
averaged time series for each event type?
I hope all this makes sense. I'm sorry but it's the very first time I
do this and I'm trying to make some sense out of all the information
I've gathered.
hope you can help
Many thanks
Elisa