Treatment response profiles: An extension of the double dissociation concept in neuropsychological research.

The presence of double dissociations in patients with neurological damage has long been used as evidence that the dissociated functions cannot be explained in terms of a common system or module. Shallice (1988) has suggested that a second procedure, the double critical variable method, can provide evidence for a similar conclusion. In this paper we examine the situation where double dissocations are not naturally present, suggesting that the two phenomena are merely aspects of the same underlying condition. We propose that the logic of the double critical variable method can be applied in this situation, whenever responses to treatment vary in a particular manner across syndromes and patients. This logic was previously used by Beschin, Cocchini, Allen, and Della Sala (2012) to show a dissociation between anosognosia and neglect in stroke patients; we suggest that it might have a more general application. As an aid in understanding the concept we also introduce the performance/performance curve; this builds on the existing idea of performance/resource curves to draw a single graph from two such curves, whose points may be derived from direct observation. It enables the empirical testing of hypotheses about the functional form of unobservable performance/resource relationships, and may be of use beyond the existing application to treatment response profiles.