[FieldTrip] Fieldtrip Discussion List

[klocke p.r.o. (pk1e14)]

Dear Fieldtrip-Community,
I was wondering what statistical test (independent-samples T or dependent-samples T) is needed to perform Cluster-based permutations in a within-UO design (e.g. 12 subjects, 2 conditions) where there are unequal counts of observations for individual subjects. For example, a clinical observation was not able to be collected at the time of recording, rendering two groups with uneven counts of observations. Would this be subject to ‘ft_statfun_indepsamplesT’ given the uneven group counts (Condition A vs. Condition B)?
As far as I am concerned, the cfg.design matrix for depsamplesT is always balanced, hence I get an error when using uneven counts. I am unsure whether an independent samples-T test is the correct choice here.

Many thanks,
Philipp
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20230814/70c20a08/attachment.htm>