Santo F. Camilleri

An Application of the Shevky Social Area Indexes to a Model of Urban Society

The equations of model (4) were also estimated by the regular least-squares technique. The following parameters, together with the standard errors of regression coefficients, were found as shown in equation below. The levels of significance of the parameters are not high: a12 and b23 are significant at the 0.01 level, while bi1, a2i, b32, and bm are significant at the 0.05 level. All other parameter estimates failed to meet even the 0.05 requirement. The correlation coefficients, the standard errors of estimate, and the levels of significance are given in the following table:

An Investigation of the Utility of Urban Typology

theory of crime and criminal behavior, given the present state of sociological and psychological theory. The comparative research reported here constitutes a first step in the necessarily complex task of formulating and testing special etiological principles for specific types of criminal behavior. The empirical data, both quantitative and qualitative, demonstrate clearly that the armed robbers differ in kind and degree from other criminal types in terms of theoretically relevant social and psychological background factors.

Further Comments on the Utility of Urban Typology

approach grew out of a wide variety of high level concepts. It developed from theories of social change, economic development, social organization, social differentiation, and other macroscopic aspects of societies. It includes some rules of application for some major concepts, indicators having considerable generality and power, as earlier work by Van Arsdol, Camilleri, and Schmid demonstrates. These concepts make more sense in harness than they do separately; this is an indication of potential unification in the social science map of complex society and its transformations. However, they are by no means an exhaustive set of concepts which are useful in dealing with urban society, nor are all of the rules of application clear. Had Van Arsdol, Camilleri, and Schmid concentrated their work upon analysis of problems at this level, they might well have contributed significant new leads in the development of an admittedly imperfect (but useful and growing) approach to modern society. Instead, they assumed a logical completeness in the system (and assumed they understood it) and self-evident rules of application. When, in the process of analysis, these assumptions broke down, they shifted their level of operation from the test of congruence between theory and observation to the low order empirical-descriptive, the hunting and gathering methodology of those satisfied to find any hanging-togetherness among things. Such a shift is common in sociological hypothesis-testing. The abrupt and unfounded nature of their conclusions are indicative of anomia with respect to the problems of theory construction a d testing, common among sociologists. This is perhaps predictable in a discipline which devotes thousands of man hours to the teaching of scaling, correlational analysis, sampling (in other words, the basic descriptive operations) and nearly nothing to the methodology of theory construction a d the logic of proof. In conclusion, we are not arguing that social area analysis is without defects and ready to stand as the best possible approach to the study of urban differentiation. We take, rather, the position that social area analysis represents an important theoretical synthesis and research lead which should be followed up by others. It is parsimonious, yet has wide applicability. Methodological improvements and refinements should be made when there are adequate grounds for making them; further elaboration and specification of the underlying theory is required; more empirical work relating the Shevky indexes and the social area types to theoretically meaningful dependent variables, at more different imes and places, is needed; experimentation i volving a wider variety of methods of analysis-such as exploring the relationships between contextual, unit, and individual variables-is called for. All these things are now being done in this country and elsewhere. Until more results are in, it is our belief that social area analysis constitutes a unifying, general, powerful, efficient, and sociologically-meaningful approach to the study of urban phenomena, with a utility for urban research far greater than that of any other single approach which currently exists. However, we hope that the method of social area analysis, as it now exists, will eventually be superseded by some far better system of analysis-perhaps a modified version of the current methods of social area analysis or something entirely different.