[Eeglablist] how to deal with subjects that have more components in one cluster?

[Carolina Pletti]

Dear all,

I am trying to analyze the data of an ERP study using EEGlab study and
dipfit.
I have 2 within-participant conditions and 2 groups.
In practice, what I would like to do is analyze the EEG at the component
level, in order to see:
- if there's any component that differs between condition and where in the
brain it origins
- if there's any component that differs between groups and where in the
brain it origins
I am also interested in interactions effects between group and conditions.

Now, I did the dipole estimation for each participant with Dipfit, I
created a Study, and then I clustered the components based on dipoles,
spectra and ERPs. The results look good, but some participants have 2 or
three components clustered in the same cluster. How do I deal with that in
the statistical analyses? Isn't it a problem if a participant is sampled
more often than others, especially if data are considered independent like
in between-subject comparisons?
Also, some participants don't have components for some clusters. Is this
taken into account in the analyses of the between-participant effects, or
are the test just run on the participants that are included in the clusters?

Thank you,

Carolina
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20180309/b9de2f20/attachment.html>