 |
 |

Nonmetric Pattern Analysis of Behavioral and Biological Disease Symptoms in Alcoholism
Thomas A. Layden, MA;
James W. Smith, MD
Arch Gen Psychiatry. 1973;28(2):246-249.
Abstract
In order to present an empirically derived two-dimensional representation of alcoholic symptomology, correlations among 30 behavioral and biological alcoholism variables were subjected to Guttman-Lingoes nonmetric factor analysis and monotone distance analysis. The factor analysis provided a relatively distortion-free solution with ten factors and the distance analysis provided a two-dimensional graphline representation in which six of the ten factors were easily discernible. Contiguous regions in the graphical representation depicted four salient aspects of alcoholic symptomatology: (1) emotional state variables, (2) enzyme variables, (3) perceptual motor and intellectual variables, and (4) liver function variables and age. These results suggested a progression of disease symptoms in which increased enzyme activity might not be accompanied by intellectual de terioration, but in which such deterioration would be a probable concomitant of advanced biological disease.
Author Affiliations
Seattle
From Shadel Hospital, Seattle. Mr. Layden is now with ICI America, Inc., Wilmington, Del.
Footnotes
Accepted for publication July 3, 1972.
Reprint requests to ICI America Inc, Wilmington, Del 19899 (Mr. Layden).
CiteULike Connotea Del.icio.us Digg Reddit Technorati Twitter
What's this?
|