Ok, thanks for clarifying! I think I get it!
I estimated the single trial sources as you said but then averaged the single trial source estimates for each participant. I then use the subject average to compare my subjects (in SPSS). So this is one processing step in addition that might not have been necessary (thanks, so I know this for next time!) but that should not change my results if I get you right?
Sorry for all the confusion!
I only use the group average for display purpose so it should be ok.
Thanks
Maren