ttest that accounts for the number of trials in each condition of interest

Is there a ttest-like program that can take into account how many trials (or how few!) one has for each condition of interest?


Hi Christine,

Perhaps Gang can correct me, but 3dMEMA might do that to some degree. Subjects with fewer events would probably explain less variance, and therefore be weighted somewhat less than subjects with more events when computing the group beta.

  • rick