Dimensions of personality, psychiatric syndromes, and mathematical models.

Abstract

It is well known that psychiatric diagnostic groupings have no claim to represent any fundamental scientific causal principle, but reflect rather the needs of administrative convenience and compromise between different theoretical orientations (Eysenck, 1960). Doubt has in fact been expressed regarding the advisability of retaining categorical divisions in a field where quantitative differences along orthogonal dimensions may be more appropriate than qualitative differences between distinct disease groups (Eysenck, 1947). The appropriate statistical method for dimensional analysis is, of course, multiple factor analysis (Eysenck, 1952) and it is possible to show relationships between factors or dimensions and psychiatric categories by giving factor scores to the subjects of the experiment and to average these scores for groups of subjects sharing a common diagnostic label (Eysenck, 1959). In this way it has been demonstrated that along the dimension of extraversion/introversion, subjects diagnosed as psychopaths tend to have particularly high extraversion scores; hysterics tend to be extraverted but not as highly as psychopaths. Patients suffering from one of the dysthymic conditions (anxiety, reactive depression, obsessional disorders), tend to have high scores on introversion. Mixed neurotics tend to be in between the other groups. All these diagnostic groups have high scores on the factor of neuroticism which is orthogonal to extraversion/introversion (Eysenck, 1957).