probability zero

[Eric Maris]

Hi Giovanni,


It is good to know that you would not have this question if the support of
the permutation distribution was continuous instead of discrete, because in
that case >= and > would produce the same p-value. In most situations, the
number of replications is so large that the difference between >= and > is
very small. I am aware that this answer is not acceptable for someone who
appreciates the discrete nature of the support of the permutation
distribution. To those neuroscientists, I can only say that whether you use
>= or > is a matter of choice. This choice is of the same type as choosing
for 0.05 or 0.01 as your significance level.

Best,

Eric



-----Original Message-----
From: FieldTrip discussion list [mailto:FIELDTRIP at NIC.SURFNET.NL] On Behalf
Of Giovanni Piantoni
Sent: maandag 8 februari 2010 15:38
To: FIELDTRIP at NIC.SURFNET.NL
Subject: [FIELDTRIP] probability zero

Dear FieldTrip users,

I am using statistics_montecarlo, which works really well, and I have a
theoretical question. After using clustering correction for multiple
comparisons, some clusters have a probability of 0.
I don't know how to interpret this zero, as the lowest p-value you can get
with a nonparametric method is 1/number_of_randomizations (see Nichols &
Holmes, 2002).

Shouldn't the observed cluster statistic be included in the reference
probability distribution?
Or has '>=' (greater than or equal to) maybe become a '>' (greater than)?

Thanks,

Giovanni

------
Giovanni Piantoni, Ph.D. student
Dept. Sleep & Cognition
Netherlands Institute for Neuroscience
Meibergdreef 47
1105 BA Amsterdam (NL)

+31 (0)20 5665492
g.piantoni at nin.knaw.nl
www.nin.knaw.nl/research_groups/van_someren_group/

----------------------------------
The aim of this list is to facilitate the discussion between users of the
FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG
and EEG analysis. See also
http://listserv.surfnet.nl/archives/fieldtrip.html and
http://www.ru.nl/neuroimaging/fieldtrip.

----------------------------------
The aim of this list is to facilitate the discussion between users of the FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/neuroimaging/fieldtrip.