A Comparison of Indexes of Socio-Economic Status

IN the past twenty years many indexes have been devised as measures of the of individuals. The task has been complicated by the fact that investigators have not agreed upon precise definitions of the term socio-economic status (or its several synonyms); therefore they sought a short and usable index for a loosely defined and unmeasured variable.' This variable has been conceived of in different ways: as a unidimensional attribute that could be directly measured if we had adequate tools; as a unidimensional attribute, but one that must be measured indirectly; as a unidimensional composite that cannot be directly measured, made up of several interrelated attributes that are measurable and can be combined in an index; as a complex of attributes that are interrelated, but do not form a single dimension and thus should not be measured, directly or indirectly, as a totality. Many researchers have avoided the logical and definitional problems by using a measurement that supposedly maximized prediction of certain behavioral consequences of position, however it be defined. Usually these investigators have offered a new index that proved useful in the context of a given research situation without indicating how it was related to other measurement devices. The result has been proliferation and confusion, with only a few attempts to study the relations among the indexes themselves.2 This article reports an additional study of the inter-correlations among standard measurement tools. Information was collected from over two hundred adult men; 19 scores were computed for each respondent which measured some aspect of his status; these scores were then inter-correlated and subjected to factor and cluster analysis. The 19 measures were not all independent; many were simply alternative ways of scoring the same variable (such as occupation), for the field worker needs guidance about their relative efficiency. Some additional data are presented comparing open-ended and closed questions on the self-identification of the respondents.