Figure 10.

Group-level motion control strategies have a minimal impact on results. Including motion in the group-level design matrix (middle column) or using motion-matched samples (right column; see Fig. 2) produced nearly identical results. Age-related effects were only minimally stronger when motion was not accounted for at the group level (left column). In all cases, brain activation to WM load was much more strongly related to performance than age. The right lateral cortical surface is displayed; similar effects were seen elsewhere.