Using Independent Components Analysis to Better Understand the Electrophysiology of Error Processing

Abstract

The Pe (error positivity) is a positive deflection in the electroencephalogram (EEG) peaking around 250ms following a response when a person makes a task error, especially if they are immediately aware of having made that mistake. In our two prior studies of error processing, participants completed a 2-choice response task (the Flanker task) to invoke error-related EEG activity, including the Pe. The Pe normally appears as a positive deflection at posterior electrode sites, and we observed this, but we also saw what appears to be a mirror-image negativity at frontal sites at the same time; we tentatively call this an “inverted Pe”). It is unclear, however, whether these two deflections reflect the same underlying neural activity or separate neurocognitive events. We are currently using independent components analysis (ICA), a computational method for separating complex, mixed signals into simpler subcomponents, to address this question. We predict that the canonical Pe and “inverted Pe” will be appear together in ICA components, suggesting a common neural generator. If they appear in separate ICA components, that will imply distinct neural generators. In either case, resolving this question will advance our knowledge of the electrophysiology of error processing.