[Eeglablist] statcond

[Arnaud Delorme]

Dear Alexander,

> Now I'm wondering if the p-values that
> are returned from this function are already corrected for multiple
> comparisons or not?

No, the p-values are not corrected for multiple comparisons yet. One way 
to do it is to use FDR (the webpage below contains a Matlab function).

http://www.sph.umich.edu/~nichols/FDR/

Another elegant (but computationally expensive) way to do it is to 
bootstrap the size of the biggest cluster of contiguous significant 
values as this is performed in Fieldtrip.

We will have both of these soon in EEGLAB.
Also, we figured out that bootstrap was actually better than 
permutations (for reasons pertaining to generalization of hypothesis 
from the data sample to the population from which it is drawn). The 
latest statcond function will thus allow to use bootstrap.
Best,

Arno