What is an ERP?
A basic introduction to electroencephalography and event-related potentials.
![]()
The Event Related Potential (ERP) technique allows us to take raw EEG
data, the electrical activity recorded from the brain, and use it to investigate cognitive processing. First, we record a
subject's EEG while they perform a task designed to elicit the proper
cognitive response (eg, attending to a certain type of object). To
accomplish this, we have subjects wear a mesh cap embedded with
electrodes which record brain activity. We also attach electrodes to
the face to monitor eye movements.

(Figure 1)
The raw EEG recorded directly from a subject (Figure 1) is a summation of all the electrical activity occuring in the brain at a given moment in time. In this form, it tells us little more than that the subject is awake or asleep. To parse out only the information we are interested in, we must time-lock and average the signal. This means that we mark the point in time when a stimulus occurs and average together a large number of trials so that everything that happens 100ms poststimulus, for example, is averaged with everything else that occurs 100ms poststimulus. (Figure 2)

(Figure 2)
This averaging process allows us to filter out all brain activity that is not related to the appearance of the stimulus. Any neural activity unrelated to the experiment will not occur at a consistent time relative to the stimuli. This random variation will be averaged out given enough trials. This leaves us with an ERP waveform isolating the electrophysical activity related to that visual stimulus. Many of the waveform's peaks have been linked to specific cognitive mechanisms. Because of the electrophysiology involved, ERP's have a very fine temporal resoloution but little spatial resolution. Temporal information, however, is fundamental to understanding how something works. Thus, ERP's are a crucial tool in understanding cogntive processing.