Abstract
Ever since Durkheim's classic study Suicide, sociologists have grappled with the problem of integrating analyses of individual behaviour and the social contexts within which that behaviour occurs. I review some of these issues to show how recent developments in multilevel models provide not only an example of convergence between theory and method but also a partial solution to this long-standing problem. I use Australian data on ethnic marriage patterns to illustrate how relevant features of the group context can be introduced into empirical analyses in order to understand variations in ethnic homogamy between groups and over time.
Get full access to this article
View all access options for this article.
