We argue that if a typological system is to be more than an efficient means of communication, if it is to have predictive and explanatory power superior to approaches that make no assumptions about categories, the existence of discontinuities in the relevant data must be established empirically. Two forms of discontinuity are discussed: (a) discontinuities in the univariate or joint distributions of the variables that define the types and (b) discontinuities (configurality) in relationships with variables external to the system. It has been suggested that the Socialization and Femininity scales of the California Psychological Inventory can be combined to yield distinct types of individuals, but analyses of two data sets failed to reveal either form of discontinuity. Rather, the data are best understood as reflecting a linear additive combination of two continuous dimensions. At present there, does not seem to be any typology in personality research that is demonstrably more than a simplifying way of talking about complex, continuous data.We wish to thank Jack and Jeanne Block and Harrison G, Gough for their invaluable contributions to this research. They not only made available to us the data analyzed and reanalyzed in this paper but also provided us with many suggestions, criticisms, and observations that helped us to sharpen our thinking. We also wish to thank the students in the seminar in personality research where this project originated. They helped to raise questions about typologies and to do several of our first data analyses.

No abstract

No abstract

No abstract

No abstract