SMABS 2004 Jena University
SMABS 2004 Home Organization About Jena Sponsors Links Imprint / Contact SMABS Home

European Association of Methodology

Department of methodology and evaluation research

Jena University

Contributions: Abstract

The analysis of social cleavages: distance- versus attribute models

Georg Mueller
University of Fribourg
Switzerland

A social cleavage is a division of a human society into groups, which are characterized by external antagonisms with regard to values, behaviors, or structural attributes and internal similarity of the same criteria. The separation of societies by ethnicity, language, or class, are typical examples of such cleavages, which are traditionally analyzed by looking at the individual attributes of the cleavage group members. A more natural way of investigating social cleavages is however the analysis of the distances between individuals in a multidimensional attribute space: Pairs of individuals belonging to the same cleavage group should on certain dimensions be closer together than other pairs which do not belong to same group.

Since primary data such as the records of standardized interviews generally do not contain information about distances between individuals, the paper proposes to gather this information by randomly matching individual data records. The units of analysis of the resulting "artificial" dataset are no longer individuals but dyads of persons for which the inter-individual distances in the aforementioned attribute space can easily be computed. In the subsequent statistical analysis, these distances are used in order to determine by logistic regression and loglinear modeling the main dimensions of a cleavage as well as the degree of mutual antagonism and internal homogeneity of the analyzed cleavage groups. Since the metric of the distances can be adapted to the available level of scaling and since the number of possible dyads of analysis is the square of the size N of the original sample, the proposed method is relatively "robust". This advantage is demonstrated by an exemplary analysis of data from the International Social Survey Programme, which refer to value cleavages about gender roles.