[FieldTrip] grandaveraging EEG question

[Dave Britton]

I have EEG data that consist of two separate file sets for each subject, 
and 20 subjects. When I process this to get ERPs for various conditions, 
does it matter if I just use ft_appenddata to concatenate all 40 files 
together and then run ft_timelockanalysis on the result, or should I use 
ft_grandaverage to combine the two runs for each subject, using the 
"within" parameter" and then combine the 20 resulting grandaverages with 
another call to ft_grandaverage using the "between parameter?
-Dave