Figure 6.

Establishing subject consistency and activation significance. a, Consistency of activations for each of the three components across subjects. Colored voxel represent the number of subjects that showed significant activations. A voxel value of 12 indicates that voxel was activated in all subjects in our dataset. b, Unthresholded z-map for a permutation test where we randomly relabeled all trials, remodeled the fMRI data based on the principles of Figure 2 and repeated the group analysis. The original activations for the early, difficulty and late components, the boundaries of which are depicted by the green outlines, are no longer present (no other activations survive after correction for multiple comparisons either). This negative result reinforces the significance of the original effects which appear to only correlate with the “true” trial labels.