Measuring endogamy

Abstract

This paper is concerned with the description and statistical analysis of marriage patterns among a set of intermarrying groups. The statistic K is suggested as a reasonable measure to summarize the overall tendency to endogamy, and the measures proposed by Rommey (1971; in Explorations in Mathematical Anthropology, M.I.T. Press, Cambridge) in his pioneering paper are critically evaluated. The marriages recorded in the data occur sequentially, and the probability of each might be assumed to depend only on the group membership of the two mates. These marriage probabilities, which are indicators of the preferences for different marriages, can be estimated simply if the data are assumed to be a sample from a large population. If we ask questions such as whether the tendency to endogamy is equally strong for each group, or whether there are preferences for some exogamous marriages over others, summary descriptive endogamy measures will not suffice. We describe a class of statistical models of great flexibility which permits us to test such hypotheses and measure the degree of departure of the data from them. Finally, the statistical procedures proposed in the paper are applied to some sets of empirical data from the literature, both to illustrate the techniques and to indicate what conclusions should be drawn in these instances.