Reliability and the discrimination of normal and pathological groups.

The authors of a recent article (Neufeld, R. W. J., and Broga, M. I. Evaluation of information sequential aspects of schizophrenic performance. II. Research strategies and methodological issues. J. Nerv. Ment. Dis., 169: 569-579, 1981) questioned the usefulness of considering reliability and other psychometric characteristics of tasks in the comparison of the pathological and normal groups on the abilities measured by the tasks. They supported their conclusion by a reanalysis of data published by Chapman and Chapman (Chapman, L.J., and Chapman, J.P. Problems in the measurement of cognitive deficit. Psychol. Bull., 79: 380-385, 1973). We contend that their conclusion was based on an inappropriate analysis of the data with a neglect of relevant psychometric principles.