If done for each point, same test repeated and false positives increase
eg: t test in each electrode
A result is more believable if it occurs in a set of adjacent channels:
Threshold data of statistical test
Compute clusters
Threshold randomized data
Compute clusters for 100 or so distr of randomized data
Decide if its rare
sumT=sum of all t stats
There was a significant difference between easy and more difficult trials between 712 ms post-stimulus and 768 ms post-stimulus. This difference was initially localized to a few central electrodes but over time spread out more posteriorly. This is consistent with previous studies that have shown
