afni_nonparametric_options

[afni_user_xiy]

Dear Afni experts,

In addition to using 3dMVM in the 2nd level analysis for a two-way mixed design, I wish to continue with an “equivalent” nonparametric test. Do we have something similar to SPM’s SnPM and FSL’s randomize?
My understanding is that with afni, we can use 3dClustSim to correct the cluster threshold and I also read a bit about afni BML… I would really appreciate any suggestions on some up-to-date afni nonparametric approaches.
http://andysbrainblog.blogspot.com/2017/01/problems-with-and-solutions-to-cluster.html

Thank you!
Xi

[pmolfese]

There are some options within AFNI for nonparametric statistics, see here.

And the bayes approaches can be found here[/url], [url=https://afni.nimh.nih.gov/pub/dist/doc/htmldoc/tutorials/meta/basic_bml.html]here[/url], and [url=https://afni.nimh.nih.gov/pub/dist/doc/htmldoc/codex/fmri/2018b_ChenEtal.html#]here. Although i may have missed a link or two.

Is there a specific thing you’re hoping to do with your data that requires non-parametric statistics?

[ptaylor]

Just adding a brief addition to useful Pete’s list:

• Equitable Thresholding and Clustering: A Novel Method for Functional Magnetic Resonance Imaging Clustering in AFNI
  https://pubmed.ncbi.nlm.nih.gov/31115252/
  (for ETAC)
• FMRI Clustering in AFNI: False-Positive Rates Redux
  https://pubmed.ncbi.nlm.nih.gov/28398812/
  (“3dttest++ -Clustsim …”)

–pt

[afni_user_xiy]

Thank you both for the quick responses and extremely helpful resources!
Xi