[FieldTrip] One-sample T-test with cluster-based statistics at the group-level

[BROVELLI Andrea]

JM you are the best! Since 2004 :)

Strangely enough I could find this solution on the discussion list. Thanks a lot. We'll try this out.

Andrea
________________________________
De : fieldtrip <fieldtrip-bounces at science.ru.nl> de la part de Schoffelen, J.M. (Jan Mathijs) <jan.schoffelen at donders.ru.nl>
Envoyé : samedi 11 mai 2019 12:01:59
À : FieldTrip discussion list
Objet : Re: [FieldTrip] One-sample T-test with cluster-based statistics at the group-level

Caro Andrea,

Yes, it can be done.
The procedure would be to duplicate your data object, and to replace all functional values with zeros, then do the statistical inference with cfg.statistic = ‘depsamplesT’; This has been possible since 2004 or so, which was before the existence of MNE-python, as you are probably aware :).

Note that it is indeed (as mentioned in your quote) statistically questionable to test a single condition against 0 under the permutation framework, since there are no conditions to permute.

Alla prossima,

JM


On 11 May 2019, at 11:22, Andrea Brovelli <andrea.brovelli at univ-amu.fr<mailto:andrea.brovelli at univ-amu.fr>> wrote:


Dear all,

I'm sorry to insist on this issue which has largely discussed on the mailing list (https://mailman.science.ru.nl/pipermail/fieldtrip/2018-October/025301.html) (https://mailman.science.ru.nl/pipermail/fieldtrip/2018-August/012314.html), but I cannot find an easy way out to perform a one-sample t-test with cluster-based statis at the group (or single subject) level using Fieldtrip.

As you are probably aware, MNE-python has implemented such test in this code<https://martinos.org/mne/stable/generated/mne.stats.permutation_cluster_1samp_test.html> starting from the fact that [lines 1182-1186]:

"Because a 1-sample t-test on the difference in observations is mathematically equivalent to a paired t-test, internally this function computes a 1-sample t-test (by default) and uses sign flipping (always) to perform permutations. This might not be suitable for the case where there is truly a single observation under test"

Is there a way to "hack" the current version of Fieldtrip to perform such test?

Are you planning to implement it in the near future?

Thanks a lot.

Cheers,

Andrea

_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20190511/905cbd2c/attachment-0002.html>
-------------- next part --------------
_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202