Dear FSL experts,
I have draw a ROI-box of 10 mm size....I have used as a mask (-m) in randomise....I have tried with 5000 permutations (independent t-test) and I have almost reached significant results tfce corrected (0.92 for on group vs 0.02 for the other group). Now, I am trying the same but with 10000 permutations in order to get a more adjusted distribution of my data.
The question is: could the size of my box affect the results of my statistics?, that is to say: my results could improve if the size of my box was, i.e 20 instead of 10?
Any suggestion as well as explanation would be highly appreciate.