Yes, and here there also seems to be problem with what they count as “activity” in a given region of the brain. An fMRI produces a huge amount of data that needs interpretation, and so is ripe for statistical fudging.