Hello,
Since a cluster doesn’t have a single t-statistic ( its statistic is its extent, which can’t really be expressed as a simple correlation ). You could take the underlying t-statistic image that the cluster was formed from and calculate a “summary” partial correlation using ( e.g. ) the peak t-statistic in the cluster.
Hope this helps,
Kind Regards
Matthew
> On 5 Jul 2016, at 15:26, Shady Rahayel <[log in to unmask]> wrote:
>
> Hi,
>
> Is there a way to have a coefficient correlation (like Pearson's or Spearman's) when one has significant clusters? For the moment, I've got clusters at p<0.05 corrected but I'm wondering whether there is a way to really indicate the "strength" of that relationship?
>
> Thank you,
>
> Shady
|