The relation between neurosis and psychosis: a reanalysis.

RECENTLY D. S.Trouton and A. E. Maxwell (7)reportedtheresultsofa factorial analysis of 45 variables derived from an Item Sheet used at the Bethlem Royal and Maudsley Hospitals. Their sample consisted of 819 male hospitalized patients aged 16—59having non-organic psychiatric disorders. The variables included the major symptoms and past history notations of psychotic and neurotic patients. Their analysis yielded six factors, the first two of which were identified as “¿ psychoticism― and “¿ neuroticism―. The authors conclude that their findings support the view that the varied symptom patterns of mental illness can be accounted for by these independent components. An examination of the two-dimensional plots of the orthogonally rotated factors indicated that a much better fit to the data could be obtained if the axes were oblique rather than orthogonal. On apriori grounds it would seem plausible to expect that the underlying dimensions isolated in this domain would be correlated. Moreover, there is considerable supporting empirical evidence for this belief in the studies of psychotic and neurotic symptoms reported by Degan (1), Lorr, Jenkins and O'Connor (3), and Lorr and Rubinstein (4). The goal of the Trouton and Maxwell rotations was to find a transformation “¿ which would yield factors conforming as far as possible to present psycho logical knowledge and psychiatric practice and at the same time aim at economy of description―. The purpose of this study was (a) to re-rotate the original data to a position that would more closely approximate simple structure, and (b) to suggest alternative interpretations of the rotated factors.