[Eeglablist] dB / mV inconsistencies

[Chrissy Lackner]

EEGlab list,
I have inherited a set of longitudinal EEG data from a colleague, and  
it seems as though there was an error made while recording the data  
during one of the two time points for one participant. The dB and uV  
are WAY off for time one. They are not even comparable to the uV for  
time two, or comparable to any other subject at time 1. It does look  
like there is meaningful data underlying all of this, and I am hoping  
someone can help me figure out a way to extract all of this data so I  
don't have to loose this participant.

I've uploaded here  (http://db.tt/fqCZK0xM OR here  
http://db.tt/sXxiYREU)  a screenshot of their time one data (note the  
scale on the bottom right and the odd TF decomposition) as well as  
their time two data (normal).

I have tried re-referencing to average, refiltering, removing the  
'real' channel mean as discussed in a previous list message:

> % remove *real* channel mean
>      for numChans = 1:size(EEG.data,1);
>      EEG.data(numChans, :) = single(double(EEG.data(numChans, :)) -
> mean(double(EEG.data(numChans, :))));
>     end

And finally I've tried normalizing (zscoring the data). This last   
approach seems promising, although it means that I'd have to   
re-process all other participants in this manner. Perhaps there is a  
way to scale down the uV in a way that I haven't come across?

  Any ideas?
  Thanks for your help.
  Chrissy



  Chrissy Lackner, MSc
  PhD Candidate, Behavioural Neuroscience
  Brock University | Psychology Department
  Niagara Region   |  500 Glenridge Ave.  |  St. Catharines, ON  L2S 3A1
  brocku.ca | T  905 688 5550  x5511