Statistics
from 3dDeconvolve
¥ An F-statistic measures significance of how much a model
component reduced the variance of
the time series data
¥ Full F measures how much the signal regressors reduced the
variance over just the baseline
regressors (sub-brick #0
below)
¥ Individual partial-model Fs measures how much
each individual signal regressor reduced data
variance over the full model with that regressor excluded (sub-bricks #19, #22, #25, and #28 below)
¥ The Coef sub-bricks are the b weights (e.g., #17, #20, #23, #26)
¥ A t-statistic sub-brick measure impact of one coefficient