[FieldTrip] Number of randomizations + 1 in clustering tests?
mariana_b_rebelo at hotmail.com
Wed Mar 5 14:37:51 CET 2014
I am computing some clustering tests on my MEG data and I don't understand the p-value returned in the field stat.posclusters(1).probIf I perform 1000 randomizations I get a p-value of 0.00799201I don't understand how this is possible, the p-value should be equal to integer/1000, so I should get either 0.0070 or 0.0080It seems that in fact I am not performing 1000 randomizations as I set in numrandomization, but 1001, because then: 8/1001 = 0.00799201
Is this really the case? Is the fieldtrip script adding another randomization to what is set by the user?
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the fieldtrip