Comparing CONN results to SPM & xjview

[quote]Dear all users, and software
developers,  
My questions: 1-
In CONN (19c), how to set cluster-level p-value and pixel-level p-value for
random field theory parametric settup? Should I defined it arbitrary or any
rational selection? for example why the default setting Voxel
p-value<.001, and cluster-level p-value<.05?
2- Why cluster size in
default setting is not an option?
3- When displaing
results in SPM12(7771), the results are totally different. How to interpret
them? 
4- when
displaing CONN results in xjview, it is againg different. xjview asks
for cluster size, and p-value for FDR correction. I don't know if it is
voxel-level or cluster-level correction. I asked the software developer, but no
answer yet.  I do not know why It does
not able to report results for two-sided ttest.


Could you please help me to know how to report my results?

Your help is really appreciated. 
Best Regards,
Samane
[/quote]