The practice of rescaling scalp-recorded event-related potentials

In a recent article the principles of and the recommended practices for rescaling scalp-recorded electrophysiological data were submitted to a comprehensive review [Urbach, T.P., Kutas, M., 2002. The intractability of scaling scalp distributions to infer neuroelectric sources. Psychophysiology 39, 791–808]. The authors argued, on both conceptual and pragmatic grounds, that the practice of rescaling be discontinued when the motivation for rescaling was to infer that at least partially non-overlapping brain regions were engaged in two different experimental conditions. This article is a response to that proposal, and the key observations are that: (1) there remain sound theoretical reasons for rescaling when the motivation for rescaling is to infer that not entirely the same cognitive processing is engaged in two conditions and (2) at least for certain classes of experimental design the pragmatic concerns raised by Urbach and Kutas are not sufficient to warrant recommending that rescaling be discontinued.