Dear Chadwick,<div><br></div><div>I appreciate you are serious about it.</div><div><br></div><div>> I assume that the source of error used in the statistical test is the<br>between subject error, unless of course the authors explicitly state<br>
this is a single-subject, in which case the source of error is the<br>between-trial error. This is convention, isn't it?</div><div><br></div><div>Yes you said it. I thought ICA-clustering is another convention since it's been used for more than 10 years. There the errors are between-IC.</div>
<div><br></div><div>> Do you have any other suggestions?</div><div><br></div><div>Just let the data self-organized and see what's there. If you don't see anything significant then I can help you.</div><div><br>
</div><div>>Thus, I have been manually editing my clusters so each cluster has</div>
<div>exactly 1 IC from each subject,</div><div>> Should I merge ICs from a subject when they are over-represented in a<br>cluster? Should I exclude subjects that are not represented in a cluster<br>and explicitly state that along with the result?</div>
<div><br></div><div>I would not do any of these. Just be confident.</div><div><br></div><div>For your information, I wrote a plugin (unreleased yet) for backprojecting cluster ICs to each individual's channel(s) (e.g. Cz) so that I can have one data per subject (Jason Palmer, a developer of AMICA, actually recommended it). If you are interested in it let me know.</div>
<div><br></div><div>Makoto</div><div><br></div><div>2013/3/4 Chadwick Boulay <span dir="ltr"><<a href="mailto:boulay@bme.bio.keio.ac.jp" target="_blank">boulay@bme.bio.keio.ac.jp</a>></span></div>
<div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Dear Makoto,<br>
<br>
I've been trying to do IC clustering but I guess I just don't get it.<br>
<div><br>
On 3/1/2013 8:42 AM, Makoto Miyakoshi wrote:<br>
><br>
> If it were inappropriate then most of our publications would be bad :-)<br>
><br>
> ICA is data driven and generates subspaces where ICs are<br>
> intradependent. It also reflects individual differences. This is like<br>
> cost for this nice method. However, what actually matter is<br>
> convention, isn't it?<br>
><br>
><br>
</div>In general, reporting a statistically significant difference in a group<br>
of subjects implies that the source of error is between-subject and each<br>
subject is represented exactly once (for a given condition). If, in a<br>
peer-reviewed article, I see a statement such as, "P300 amplitude was<br>
greater after attended stimuli than unattended stimuli (p<0.05)" then I<br>
assume that the source of error used in the statistical test is the<br>
between subject error, unless of course the authors explicitly state<br>
this is a single-subject, in which case the source of error is the<br>
between-trial error. This is convention, isn't it?<br>
<br>
Since IC clusters may, at the extreme, represent a single subject then<br>
the above statement could really mean a single-subject difference even<br>
though many subjects were measured. To me, this breaks convention and is<br>
unintuitive.<br>
<br>
Thus, I have been manually editing my clusters so each cluster has<br>
exactly 1 IC from each subject, even if sometimes that means excluding<br>
an IC that fits my model if it comes from a subject with multiple<br>
clusters that fit the model or including an IC that does not fit the<br>
model if a given subject has no clusters that fit the model. I see no<br>
other way to report my results in a manner that follows convention.<br>
Editing clusters this way is biased but I'm not sure what else I can do.<br>
<br>
Do you have any other suggestions?<br>
Should I merge ICs from a subject when they are over-represented in a<br>
cluster? Should I exclude subjects that are not represented in a cluster<br>
and explicitly state that along with the result? i.e., "Left M1 cluster<br>
mu-rhythms were significantly more desynchronized during task A compared<br>
to task B among the 6 (of 10) subjects with an IC that localized to left<br>
M1 and had similar ERSP profiles."<br>
<br>
I read the MPT paper and tried using it but I ended up with an<br>
interesting domain that contained only 1 dipole (from a set of 13<br>
subjects x 64 ICs). I can't report that. I still don't understand this<br>
method fully so I do not admit defeat yet.<br>
<br>
Thank you,<br>
Chadwick Boulay<br>
JSPS Postdoctoral Fellow<br>
Keio University, Yokohama, Japan<br>
<div><div>_______________________________________________<br>
Eeglablist page: <a href="http://sccn.ucsd.edu/eeglab/eeglabmail.html" target="_blank">http://sccn.ucsd.edu/eeglab/eeglabmail.html</a><br>
To unsubscribe, send an empty email to <a href="mailto:eeglablist-unsubscribe@sccn.ucsd.edu" target="_blank">eeglablist-unsubscribe@sccn.ucsd.edu</a><br>
For digest mode, send an email with the subject "set digest mime" to <a href="mailto:eeglablist-request@sccn.ucsd.edu" target="_blank">eeglablist-request@sccn.ucsd.edu</a><br>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br>Makoto Miyakoshi<br>JSPS Postdoctral Fellow for Research Abroad<br>Swartz Center for Computational Neuroscience<br>Institute for Neural Computation, University of California San Diego<br>
</div>