[Eeglablist] Permutation testing after wavelet analysis

[Makoto Miyakoshi]

Dear Nancy,

> 1)      I performed my wavelet analysis on single trial data; is it
common practice to also perform the permutation testing on the single trial
data? When I try to select the “Use Single Trials” option I get a message
that I will have to precompute the channel measures again and save the
single trials, so I am wondering if this is necessary.

In a common practice, we do NOT use single-trial variance from a
single-subject data (separate from the argument whether it is a good
practice or not...)

> 2)      However, is it okay to do this permutation testing comparing the
data of a healthy population and a clinical population, given that they may
have different noise characteristics? Anything I should consider in this
case?
It is okay.

Makoto



On Mon, Aug 15, 2016 at 3:34 PM, Lundin, Nancy B. <nlundin at indiana.edu>
wrote:

> Hello all,
>
>
>
> I performed a wavelet analysis in EEGlab to get both ERSP and ITC measures
> for auditory oddball data from a clinical population and healthy controls.
> I am doing non-parametric permutation testing on the output to determine
> significantly different regions in responses to target stimuli between
> healthy and clinical populations, and the same for between-group response
> to standard stimuli.
>
>
>
> In Plot Channel Measures à Stats, I selected the options to compute 1st
> and 2nd independent variable statistics, EEGlab permutation statistics,
> threshold p-value of 0.05 and an FDR correction (randomization: auto).
>
>
>
> I have a couple questions:
>
>
>
> 1)      I performed my wavelet analysis on single trial data; is it
> common practice to also perform the permutation testing on the single trial
> data? When I try to select the “Use Single Trials” option I get a message
> that I will have to precompute the channel measures again and save the
> single trials, so I am wondering if this is necessary.
>
> 2)      I have read studies that have done permutation testing on
> time-frequency data to determine significantly different regions between
> conditions within the same subjects. However, is it okay to do this
> permutation testing comparing the data of a healthy population and a
> clinical population, given that they may have different noise
> characteristics? Anything I should consider in this case?
>
>
>
> Thank you very much for any advice.
>
>
>
> Nancy
>
>
>
> *Nancy Lundin*
>
> *Doctoral Student, Clinical Psychology & Neuroscience*
>
> Indiana University Bloomington
>
> Department of Psychological & Brain Sciences
>
> 1101 East 10th St.
>
> Bloomington, IN 47405
>
>
>
> _______________________________________________
> 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/20160818/4617b42f/attachment.html>