Chapter 6: ERP Analysis¶
Continuing from Chapter 5 — Events and Epoching.
With clean epochs in hand, this chapter averages them into event-related potentials and looks at the result — the same final step covered as code in the MNE-Python track's ERP Analysis chapter.
Plotting evoked responses¶
- Plot → Plot Evoked... averages the current epochs into a single evoked waveform per channel and plots it — the GUI equivalent of
epochs.average()followed byevoked.plot(). - Plot → Plot Evoked Comparison... overlays evoked responses from two conditions (two separate epoch selections) on the same axes, the usual way to look for a difference between conditions like target vs. standard.
- Plot → Plot Evoked Topomaps... shows the scalp distribution of the evoked response at chosen time points, useful for confirming a component like a P300 or N400 peaks where you'd expect on the scalp.
Beyond plain ERPs: ERSP/ERD-ERS¶
If your question is about power changes over time rather than the averaged waveform itself, Plot → Plot ERDS Maps... and Plot → Plot ERDS Topomaps... compute event-related spectral perturbation (also known as event-related desynchronization/synchronization) — a complementary time-frequency view MNELAB supports alongside plain ERP averaging.
Exporting your results¶
File → Export writes the current dataset — raw, epoched, or evoked — out to FIF, EDF, EEGLAB, or BrainVision format, with events, annotations, and bad-channel lists saved alongside it. FIF is the natural choice if you plan to continue in MNE-Python, which is exactly what the next and final chapter covers.