Hi!

I tried to use the leave one out script which was provided in the forum before and got a question. There are three kinds of spm threshold: unc. FWE and FDR. But what if I want to use a cluster level correction here? 

Another question is that if I define the between subject contrast as Allsubjects, behavchange, age, sex [0 1 0 0], between condition contrast as Pre, post [-1 1]. Do I get the significant clusters that its functional connectivity change is correlated with behavioral change while control the effects of age and sex?

If that is true, what kind of correlation coefficient is used here? Spearman partial correlation coefficient?

Thanks a lot!
Liu