Cluster formation and diagnostic significance in psychiatric symptom evaluation

  • Authors:
  • Gilbert Kaskey;Paruchuri R. Krishnaiah;Anthony Azzari

  • Affiliations:
  • Remington Rand Univac;Remington Rand Univac;Remington Rand Univac

  • Venue:
  • AFIPS '62 (Fall) Proceedings of the December 4-6, 1962, fall joint computer conference
  • Year:
  • 1962

Quantified Score

Hi-index 0.00

Visualization

Abstract

The tremendous variability in symptom constellations associated with specific psychiatric diagnoses makes the use of statistical techniques a virtual necessity in the rigid formulation of any symptom-disease model. In addition, the large number of symptoms associated with psychiatric disorders suggest that an electronic computer can be used to considerable advantage in certain aspects of psychiatric diagnosis, e.g., in the correlation analysis between symptoms, in the determination of quantitative criteria for diagnosis, and in the evaluation of the reliability of symptom assignment by different personnel. This paper reports on the results obtained from the analysis of data on 199 subjects collected by the Childrens Unit of the Eastern Pennsylvania Psychiatric Institute. The types of information available, the method of collection, and the statistical methodology and techniques are all discussed along with the results obtained using the tools of correlation analysis and simultaneous multivariant analysis of variants.