[Eeglablist] robust ERSP statistics

[arno]

It is hard to subtract all muscle components that might interact with 
the data at high frequency (because there are a lot of them and 
sometimes it is hard to decide if some components represent muscle 
activity or not).
The best approach to this might be to study the behavior of the data 
itself at high frequency and then study the behavior of a few muscle 
components. If the muscle components do not show the same behavior as 
the one observed in the raw data at high frequency, I think it is 
convincing that what you observe in the data does not arise from muscle 
activation.

Best,

Arno

Christian-G. Benar wrote:
> dear all,
>  
> I am often working with single-subject datasets, looking for 
> gamma-band ERSP changes.
>  
> I am concerned with the fact that the ERSP measurement may reflect 
> artifacts present in a few trials; does anyone has advice on handling 
> artefact rejection in high frequency, and has anyone tried to use 
> robust statistics (M-estimator, median) as opposed to the mean power ?
>  
> cheers
>  
> Christian
>  
>  
> ------------------------------------------------------------------------
>
> _______________________________________________
> eeglablist mailing list eeglablist at sccn.ucsd.edu
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20060803/acb78b06/attachment.html>