Recovery cycles of event-related potentials in multiple detection tasks.

We examined the recovery cycle of the P3 in two different experiments, one in which subjects detected near-threshold (NT) tones and the other in which they detected suprathreshold (ST) tones presented in rapid sucession. In both experiments P3 amplitudes and latencies were decremented at ISIs of 300 msec but fully recovered by 900 msec. The N1 and P2 components elicited by ST tones showed a much more prolonged (> 7.0 sec recovery cycle. These results reveal that the P3 has a recovery cycle which closely resembles that of human decision processes, a recovery cycle far shorter than those of exogenous ERP components previously examined. Sequential changes in P3 amplitude between trials were also investigated. In both experiments P3 amplitudes were largest following signal-absent trials. This suggests that subjects may have modified their expectancies about tone delivery on a trial-by-trial basis. In these experiments the P3 wave was distinguished from the resolution of the CNV on several grounds, including differences in scalp distribution, intra-trial kinetics, effect of previous tone sequences, and distributions among the subjects. These results support the position that P3 is a neurophysiological event distinct from modulations of the CNV. Differences in the amplitude and habituation of P3s produced in the NT and ST experiments suggest that the P3 may be related to decisions which require 'controlled' stimulus processing.