Could you be more specific about this? I guess shuffling the data for both channels you "de-sync" them?

 Should it be physiological asynchrony or it does not matter? The most easiest synthetic asynchrony is:

- with fft transform one signal into frequency domain

- change phase for some frequencies or for whole frequency range. It is up to you.

- transform it back with ifft into the time domain

And now you have 2 signlas with asynchrony defined by yourselve. But I agee with Tue, you should be more specific.

That's a good idea René, however if you want to measure coherence (a measure of sync in the frequncy domain) between the two channels, changing the phase will not affect the coherence between the two channels but only the imaginary part of the complex coherenCY estimate

 That is that I am not also trully sure what Sarah precisely wants. If she wants to see the asynchrony in amplitude spectrum, she should convolve on one signal some impulse responce functon, whose frequency characteristic has maximal gain in whole frequency range of EEG signal, only gives a delay to the signal. I think that delayed Dirac delta function approxiamtion could be enough.

dirac = [0 1];

dirac = [0 0 1];

eeg_asyn = conv(eeg,dirac)

Number of zeros defines the delay. The delay should be observable in amplitude then.

If she wants asynchrony in both, she need to use the filter with non-linear phase characteristic or combine both mentioned methods.