Abstract
The distinctions between diagnostic classification of schizophrenia, the description of the characteristics and course of schizophrenia, and measurement of its severity are discussed. There is confusion among these three categories by both researchers and clinicians. The causes for this confusion are discussed, and it is noted that there are few appropriate measures of cognitive function in schizophrenics. It is proposed that the research design used in most schizophrenia research does not lend itself to the development of more informative interval scales. It is proposed that rather than using performance as a dependent measure with a task of fixed level of difficulty, a ‘threshold design’ should be used, where the environmental factors are varied to produce a given performance level.