Abstract
A long-latency comnponent of the averaged evoked potential recorded from the human scalp varied in close relationship with subjects' perceptual reports in an auditory signal detection task. Detected signals evoked potentials several times larger than did undetected signals, falsely reported signals, or correctly reported nonsignals. The threshold signal intensity at which detection perfornmance exceeded chance levels was identical with concurrently obtained electro-physiological measures of threshold.
