[FieldTrip] cluster analysis

Stephen Politzer-Ahles politzerahless at gmail.com
Sun Sep 30 15:46:09 CEST 2012


Hi Mark,

The Fieldtrip wiki has examples of published papers that report cluster
analyses. In what I've seen, usually just the *p* is reported (if even
that), along with the latency and topography of the cluster. See e.g.
Pijnacker et al. (2011, *Journal of Cognitive Neuroscience*) and Meltzer &
Braun (2011, *Frontiers in Human Neuroscience*). There may be others in
http://fieldtrip.fcdonders.nl/publications

As for your second question, my understanding is that the test statistic
doesn't matter (one of the points of the cluster test is that it works no
matter which test statistic you use (Maris & Oostenveld, 2007)). Also, to
the best of my knowledge, there are no degrees of freedom for the cluster
test, since it's non-parametric (it's not based on the distribution of a
test statistic, it's based on Monte Carlo estimation).

Best,
Steve


Message: 1
> Date: Sun, 30 Sep 2012 11:30:35 +0200
> From: Mark Noordenbos <mark.noordenbos at gmail.com>
> To: fieldtrip at science.ru.nl
> Subject: [FieldTrip] cluster analysis
> Message-ID:
>         <CAJSb8nOoV4-sUsQUdSDVeYL=
> 7qVRZp7goj3AhoE7xaq8tqTPsg at mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
>
> Dear Fieldtrippers,
>
> I was wondering how you report the results of the cluster analysis in your
> papers. I have seen different types of how the results are reported.
> For example, some report only the p-value of the cluster whereas others
> report clusters as T(degrees of freedom) = [clusterstat value], p-value.
>
> Is T (multivariate t) the correct test statisic for the cluster analysis
> (like F for anova)?
> Where can I find (or calculate) the degrees of freedom of the clusterstat
> value?
>
> Thanks
>
> Best,
> Mark
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <
> http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20120930/fe6908ab/attachment-0001.html
> >
>
> ------------------------------
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20120930/d5471f2a/attachment-0002.html>


More information about the fieldtrip mailing list