PROBLEMS OF CRITERIA IN DIAGNOSIS: A PROFILE SIMILARITY APPROACH *

R. M. Knights

DOI: https://doi.org/10.1111/j.1749-6632.1973.tb43169.x

IF: 6.499

1973-02-01

Annals of the New York Academy of Sciences

Abstract:One way of reviewing advances in a field of scientific study is to examine changes in the selection and use of criteria. The criteria used in neurological and psychological studies have changed from time to time as a function of the sensitivity of the assessment techniques as well as of the investigator’s ability to coordinate and integrate information. For example, in medicine there are rarely more than five to seven criteria for a diagnosis, because of human limitations: when looking at a specific group of symptoms, a doctor will rarely be able to recall more than a few criteria. With the development of computerized data storage and retrieval systems, however, this limitation can be eliminated. Studies of children with minimal brain dysfunction (MBD) have used a wide variety of criteria to determine their unique characteristics. For example, they have been classified on such variables as medical history, physical findings, neurological signs, biochemical data, EEG patterns, behavior ratings by parents and teachers, and scores on a large number of psychological tests. Great effort has been required in order to provide any new information in studies of this nature. In an attempt to develop novel ways of setting up and identifying new Criteria, it is now appropriate to reexamine the methods used in studies of such children.