One annoyance: it looks like their p-values weren't adjusted for multiple comparisons, which seems like a problem given the number of different comparisons they

One annoyance: it looks like their p-values weren't adjusted for multiple comparisons, which seems like a problem given the number of different comparisons they made in e.g. the fMRI results. If I'm doing my math right: if they applied the Bonferroni or Sidak corrections to the fMRI results, then only one of those comparisons would still be significant.