[Eeglablist] contrast single-condition ersp/erp with baseline/zero: statistics

[Makoto Miyakoshi]

Dear Simon,

> Regarding the data-input to statcond: so I could provide a one-element
cell containing a single data matrix with the four dimensions electrode,
frequency, time, subject?

See the statcond help. They have an example code in the help, so always
start from there since it may be a little bit tricky for the first time.

> Regarding permutation: for a one-sample test against zero, I would have
to flip the sign of the data of randomly selected subjects in order to
permute the data. As I understand it, this is not implemented. Am I wrong?

statcond({rand(1,1000}}

ruturned

NaN

So it does not run one-sample t-test I guess.

Makoto

On Fri, Sep 18, 2015 at 1:47 AM, Simon Ruch <simon.ruch at psy.unibe.ch> wrote:

> Dear Makoto
>
> Thanks for your help. I will look into this and especially into
> statcondfieldtrip(). I would like to correct for multiple electrodes using
> the fieldtrip cluster correction method and permutation based statistics.
>
> Regarding the data-input to statcond: so I could provide a one-element
> cell containing a single data matrix with the four dimensions electrode,
> frequency, time, subject?
> Regarding permutation: for a one-sample test against zero, I would have to
> flip the sign of the data of randomly selected subjects in order to permute
> the data. As I understand it, this is not implemented. Am I wrong?
>
> All the best,
> Simon
>
>
> ------------------------------
>
>
> *From:* mmiyakoshi at ucsd.edu
> *Sent:* Freitag, September 18, 2015 5:36AM
> *To:* Simon Ruch
> *Cc:* Eeglab List
> *Subject:* Re: [Eeglablist] contrast single-condition ersp/erp with
> baseline/zero: statistics
>
>
> Dear Simon,
>
> > My basic question is: what ERP/ERSP does my sound presented during sleep
> evoke?
>
> What an interesting question!
>
> > To compute statistics, eeglab requires two conditions. How can I perform
> a one-sample test against zero or the baseline?
>
> It's better to use statcond() to perform one sample T test. If you have
> trouble extracting data from EEGLAB data structure let us know.
>
> Makoto
>
> On Tue, Sep 15, 2015 at 1:25 AM, Simon Ruch <simon.ruch at psy.unibe.ch>
> wrote:
>
>> Hello
>>
>> I would like to contrast my evoked potentials/perturbations against the
>> pre-stimulus baseline, respectively against zero (for baseline corrected
>> data). I would like to perform significance tests that tell me
>> where/when/at what frequencies the signal peaks. My basic question is: what
>> ERP/ERSP does my sound presented during sleep evoke? To compute statistics,
>> eeglab requires two conditions. How can I perform a one-sample test against
>> zero or the baseline?
>>
>> My approach so far:
>> I created a dummy dataset as zero-control condition for each participant.
>> In this dataset:  EEG.data(:,:,:) = 0. I used these datasets to create a
>> repeated measures study design (stimulus vs. zero condition) and
>> precomputed study parameters. In the spectral data generated for the
>> control condition files (*.datersp), I replaced all NaNs with zeros and set
>> the baseline (chanX_erspbase) to the baseline of the real data of the
>> respective participant. This allowed me to perform tests against the
>> baseline for both ERPs and ERSPs
>>
>> While this works fine, I was wondering if there is an easier and more
>> elegant approach to what I want.
>>
>> Kind regards
>> Simon Ruch
>> --
>>
>>
>> ________________________________________________
>>
>> *Simon Ruch, Dr. phil.*
>> University of Bern
>> Department of Psychology, Division of Experimental Psychology and
>> Neuropsychology
>> Center for Cognition, Learning and Memory (CCLM)
>> Office A 362
>> Fabrikstrasse 8
>> CH-3012 Bern
>> Switzerland
>> Phone +41 (0)31 631 37 29 <%2B41%20%280%2931%20631%2037%2029>
>> mailto:simon.ruch at psy.unibe.ch <simon.ruch at psy.unibe.ch>
>> http://www.apn.psy.unibe.ch/content/team/henke/
>> http://www.cclm.unibe.ch/
>>
>> _______________________________________________
>> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
>> To unsubscribe, send an empty email to
>> <eeglablist-unsubscribe at sccn.ucsd.edu>
>> 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
>
>
>


-- 
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/20151006/8a1abbf1/attachment.html>