[FieldTrip] Enquiry into Using Cluster-based Permutation Test to Compare Differences

[u3564255 at connect.hku.hk]

Dear all,

I would like to enquire about using cluster-based permutation tests to compare differences for time-frequency power data.

I am running a mixed-design experiment where we have two conditions and two groups. I hope to first compare the two conditions within each group in terms of beta power. The second goal is to compare the difference between the two groups. To compare the difference between the two groups, I first compute the difference between the two conditions for each group, and then we perform a cluster-based permutation test to compare the difference between the two condition of one group with the difference of another group.

A code representation of the second goal is as follows:

%%%%%% For Participant Group A
> cfg = [];
> cfg.parameter    = 'powspctrm';
> cfg.operation    = '(x1-x2)/(x1+x2)';
> tfr_groupA_difference = ft_math(cfg, tfr_groupA_conditionA, tfr_groupA_conditionB);

%%%%%% For Participant Group B
> cfg = [];
> cfg.parameter    = 'powspctrm';
> cfg.operation    = '(x1-x2)/(x1+x2)';
> tfr_groupB_difference = ft_math(cfg, tfr_groupB_conditionA, tfr_groupB_conditionB);

%%%%%% Perform the cluster-based permutation test
> [stat]  = ft_timelockstatistics(cfg, tfr_groupA_difference, tfr_groupB_difference)

Could you please tell whether it is the right away to use cluster-based permutation tests to compare the two groups' differences between the conditions?

Thank you so much for your invaluable advice!

Best regards,
Samuel

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20231102/541f9a7c/attachment.htm>