Anomia and Enomia: A Methodological Evaluation of Srole's Anomia Scale

Abstract

Durkheim popularized the term anomie and stimulated a great deal of theoretical discussion and research. Subsequently, Srole constructed interview schedule items to measure personal anomia, and his items have been used by a large number of investigators. Many of these researchers implicitly assumed unidimensionality while others explicitly tested for it by Guttman scaling and factor analysis techniques. However, none of these researchers repeated Srole's latent class analysis. In this paper, representative samples of households in the Los Angeles SMSA and in a suburban city responded to the anomia items as part of larger investigations. The responses were subjected to factor analysis and Guttman scaling for a test of unidimensionality, and latent class analysis as a test of scalability and typological construction. For both samples, the responses did not meet the generally accepted criterion of a pure Guttman scale. As a result of a principal axis factor analysis only one common factor was extracted in both samples. The Green solution was used for latent class analysis and the result was the minimum one would expect from a unidimensional set of items, or two classes-the anomic and eunomic. Major conclusions are that it cannot be assumed an item, or universe of items, will foster the same results in different universes or at different times, even under somewhat similar modes of analysis. If no tests for unidimensionality and scalability are made, gradations of the Guttman scale type appear to be contraindicated and polarization appears to be more appropriate for the Srole items.