Signal-to-noise ratio sensitivity in ERPs to stimulus and task complexity:Different effects for early and late components

Traditionally, early and late ERP components art: evaluated by the latency and amplitude of their components. We investigated their signal-to-noise (S /N) ratio: how well individual ERP trials correlate with their overall aver aged ERP. High correlations indicate minimal noise or individual trial vari ability while low correlations reflect greater trial variability. Continuou s EEG activity was recorded from 57 university students during single- and dual-task conditions in a within-subjects paradigm (easy and difficult tone discrimination; with and without verbal working memory task). For early co mponents, the addition of the second task increased the correlations betwee n the individual trials and the averaged ERPs. For the late components, bot h the addition of the second task and an increase in stimulus diffi- culty decreased individual trial correlations. Results suggest that the S/N ratio analysis is a useful method fur highlighting differences between early and late components associated with stimulus processing manipulations. (C) 200 0 Academic Press.