Dear Eli,
Yes, correction on cluster level = extent (in contrast to voxel level = height). In practice, for a FWE correction on cluster/extent level you go with an (uncorrected) voxel/height threshold, e.g. .001, which defines the clusters, and then you have to look at the p values in the left column of the results table to find out whether a cluster is sig. on cluster level or not.
For the theory and implementation in SPM you could look at http://www.fil.ion.ucl.ac.uk/spm/doc/books/hbf2/pdfs/Ch14.pdf and http://www.fil.ion.ucl.ac.uk/spm/doc/books/hbf2/pdfs/Ch15.pdf , as it's a rather recent paper maybe also Woo et al. (2014, https://dx.doi.org/10.1016%2Fj.neuroimage.2013.12.058 ).
Best
Helmut
|