Single subject analysis

[Tasha Poppa]

Dear SnPM support,

Apologies if these are very basic questions- permutation testing in multi-subject scenario is clear. Yet for the single subject I'm not sure how this works, e.g., if you have only one subject, isn't the size of the exchangeability block always going to be one?

Anyway, to perform a t-test of different task conditions for a single subject, do I submit contrast files associated with each individual run, or simply the single averaged contrast estimate? Also, what does "Replications of Conditions" mean here?

Thank you.

Regards,

Tasha