thanks , to answer your questions:

1,2. we are looking at dataset produced by this command:
3dmerge -prefix bold_pval_cl3 -1dindex 0 -1tindex 0 -1clust 5.7 105 -1thresh 3.0 bold_pval+orig
where bold_pval is produced by:
3dcalc -a 'bold_decon+orig[5]' -datum float -prefix bold_pval -expr "-log10 ( fitt_t2p ( a,117 ) ) * (step(a)-step(-a)) "

3. expected results : if there are a cluster of dots that is more than 3-4 dots on 3-4 consecutive slides in bold_pval, then it should show up in the clustered results, which is bold_pval_cl3

4,5. we didn't see some dots in bold_pval_cl3 as we expected them to show up as examined the bold_pval dataset (unclustered) .

Your explaination is no doubtly correct, but we believe the cluster we saw did appear on consecutive slides.

any idea? thanks.