Dear SPM experts,
I would like to follow this question. In VBM study from SPM results, it has set-level p-value, cluster level p-value and voxel-level p-value. My question is: if I get 8 clusters from the results, the set-level p-value < 0.001, but in cluster level, only 3 clusters show the significant changes (for example, FWE < 0.05 with large spatial extension) and others are not (e.g. FWE > 0.05, but still have significant change in voxel level and fit for cluster spatial extension requirement). Should I manual delete these 5 clusters and only report the 3 clusters ? or what is the best way to do ? is there any guide to interpretate statistic results in spm ?
thanks help.
JLH
|