Measurement of maximum variability within event related potentials in schizophrenia.

One limitation of averaging individual late component event related potential (ERP) responses is that a single average ERP cannot reflect the variability of responses from epoch to epoch. In this article, we describe a method to quantify this variability and determine if any part of the overall ERP reflects a maximum variance through the use of response variance curves. We then apply this method to one disorder, schizophrenia, in which variability of information processing is hypothesized to underlie aspects of the symptomatology. Response variance curves in a group of unmedicated schizophrenic patients reveal systematic differences, maximal between 190 and 250 ms, compared with those in a group of medicated schizophrenic patients and normal control subjects.