Dear Yang,
a "conjunction null" test would correspond to regions commonly activated during the different sessions, but this does not mean they are activated to the same extent. The other way round, if regions don't show up in the conjunction this does not mean they were activated to a different extent. Similar to a non-significant test, the fact that the null hypothesis is not rejected (because there was no evidence for the alternative hypothesis) does not proof there is no difference.
One possibility might be testing for equivalence, a common procedure for drug research.
Best,
Helmut
|