Background. Cognitive deficits in schizophrenia have recently been ascribed to impaired representation and use of cognitive context. Context is defined as relevant information held temporarily in mind to mediate appropriate but often non-habitual responses.

Methods. Parallel studies in a variety of cognitive domains were designed in order to explore the generality of any schizophrenic deficit in context use. In all of the tasks (a Stroop task, a Continuous Performance Task and a cued spatial location task), we examined how performance was affected by the time for which contextual information must be held in mind, and by whether context or task demands were consistent or varying between trials. It was predicted that manipulation of these variables would produce tests especially sensitive to schizophrenic attentional problems.

Results. Predictions were partially confirmed. Although increasing contextual demands failed in most cases to produce disproportionate slowing of performance in patients, error data were largely in line with predictions. At the same time, the data did not suggest a simple unitary context deficit. Instead, different aspects of context – the time over which contextual information must be held in mind and the consistency of context – were differentially important in different tasks.

Conclusions. The cognitive impairments of schizophrenic patients cannot be simply characterized as a generalized context deficit. A more differentiated, if not task specific, picture of schizophrenic deficits is suggested.