[Event-related potentials in psychiatry--methodologic principles]

The methodological requirements for the use of event-related brain potentials (ERPs) in psychiatry are critically reviewed. In particular the development of psychophysiological models is considered as essential for a functional interpretation of ERP-experiments. In addition problems of signal-analysis and clinical diagnosis relevant to ERP-research in psychiatry are briefly discussed. The logic of ERP-research in psychiatry is illustrated by a series of dichotic listening experiments.