¥ Null Hypothesis: H0 : m1 = m2 = É = mr
H i.e., subject
type has no effect on mean signal in this voxel
HAlternative Hypothesis: Ha :
not all mi are equal
H i.e., at least one subject type had a different mean FMRI
signal
¥3dANOVA is effectively a generalization of the unpaired t-test to
multiple columns of data (a further
refinement will be introduced with 3dANOVA3)
¥As
such, 3dANOVA is probably not appropriate when comparing results of
different tasks on the same subjects (need a
generalization of the paired t-test: 3dANOVA2)
¥Difference
from doing unpaired t-tests on pairs of columns: variance estimates are all pooled together, increasing the denominatorial degrees
of freedom
¥Assumption is that data
fluctuations in each column have same variance