Selavco: A method to deal with trial-to-trial variability of evoked potentials

For the identification of event-related poten-tials (ERPs) of the brain, repeated stimuli have to be presented because of the poor signal-to-noise (S /N) ratio which makes the statistical analysis of single potentials difficult. Averaging sweeps of brain activity, time-locked to the stimuli, improves the S /N power ratio in proportion to the number of sweeps averaged. A crucial assumption implicit when averaging is that identical stimuli lead to identical potentials which is a questionable and rarely verified assumption. It is well known that an experimental modification of attention can lead to a different average potential (Garcia-Austt et al. 1964; Haider 1967; KopeU et al. 1969). A conjec-ture that fluctuations of the state of attention (an