[Eeglablist] STUDY Bootstrap statistical analysis

[Makoto Miyakoshi]

Dear Jodie,

I assume you use statcond(data, 'method', 'bootstrap').
The bootstrap process is done by surrogdistrib(). It randomizes the indices
for conditions to construct surrogate data (i.e. null hypothesis that the
difference of the labels has no effect on data). Then it jumps to
ttest_cell() to run a standard t-test to test on both real data difference
and surrogate data difference separately to obtain t-values. Finally it
runs stat_surrogate_pvals() using these data to determine the significant
data point/pixel/voxel in a non-parametric way (i.e. counting from the end
to determine X-percentile).

Makoto

On Thu, Aug 13, 2015 at 5:31 AM, Jodie Feil <Jodie.Feil at weizmann.ac.il>
wrote:

> Dear All,
>
> I was wondering if you could tell me how the bootstrap statistical
> analysis works in the EEGLAB STUDY design?
>
> If I want to compare two groups (GO = 250 trials) vs (NOGO - 30 trials)
> across healthy controls and schizophrenia. I am interested in making sure
> that the differences between GO and NOGO within subjects is reliable. So I
> assumed to run the bootstrap statistical analysis (to account for the
> different number of trials). However, I can't seem to find documentation of
> what exactly the bootstrap statistical analysis is outputting.
>
> For example, when we applied the bootstrap statistic, we found a
> significance difference (p=0.01, bonferroni corrected) across a number of
> electrodes between the GO and NOGO conditions in healthy controls.
>
> I have looked through:
> https://sccn.ucsd.edu/svn/software/eeglab/functions/studyfunc/pop_statparams.m
>
> However, I would be happy if somebody could elaborate and further explain.
>
> Thank you
> Jodie
>
> ---------------------
>
> Dr. Jodie Naim-Feil
>
> Post-doctoral Fellow
>
> Research group of Prof. Elisha Moses
>
> Department of Physics of Complex Systems
>
> Weizmann Institute of Science
>
> 76100 Rehovot, Israel
>
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to
> eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to
> eeglablist-request at sccn.ucsd.edu
>



-- 
Makoto Miyakoshi
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20150817/631e7ce2/attachment.html>