1) Run art.m program on the data from msit/run1/011 (Interference Task)
 
2) Note there are lot of outliers with default thresholds however you can see that nothing exceeds 1mm or 1 rad (CHANGE THRESHOLDS)
 
3) Check “motion-task correlation” checkbox in the lower left part of the GUI
 
4) Select the SPM.mat from NOmotion directory The Color Bar reflecting amount of SCM, the colors represent the correlation coef between the task conditions (rows) and the motion parameters (columns).
 
Note: As we’ve seen in Johnstone paper, even a moderate amount of SCM (r = .2) can cause significant decrement in statistics. This illustrates the notion that you can not have a lot of absolute motion yet still have a lot of stimulus correlated motion. See SPM results with motion regressed and 2) without motion regressed.
 
Susan Whitfield-Gabrieli MAR09