[Eeglablist] Difference-Wave statistics

[Manuel Lemos Rodrigues]

Dear EEGLABlist members, 


I'm currently working on some oddball paradigm datasets and at this point analyzing difference-waves obtained by subtracting Grand-average ERPs of two conditions.
So here is my question: Which way for statistical analysis do you recommend getting time ranges where the difference waves ( n = 10-15) are significant non-zero? I'm currently using two-sided subsequent ttests for all  datapoints ( Matlab function ttest). 
I noticed that  pop_comperp does have similiar funtionality for grand-averaged ERPs with one condition, but I get different results ( more conservative results). 


Thanks a lot and best regards,

Manuel Lemos