Hi Alex,

To obtain p-values for the clusters we need to do a non-parametric test, and for that, randomise and PALM need to have the original data and model.

So, the answer is no, it isn't possible to obtain p-values for cluster mass just from the effect size maps.

All the best,


On Tuesday, 22 August 2017, Alexander S Hatoum <[log in to unmask]> wrote:

I am attempting to conduct cluster mass analysis with estimates of effect sizes from a model run outside FSL/PALM.  I have a file that has vertex-wise estimates for the effect size of each of these analyses, analogous to the effect size for each vertex in a series of GLMs.  I was wondering if FSL/PALM has support for conducting cluster mass with effect sizes previously estimated?