Ethnic groups differ on a wide range of features – spanning health, education, mental disorder, and crime, to mention but a few. These differences have given rise to a substantial literature on supposed cultural influences. In recent years, interest in cultural differences has been associated with a major paradigm shift from cross-cultural to within-culture studies (Shweder et al., 1998). For the most part, this shift reflects a concern that the concepts and measures derived in any one culture cannot be assumed to cover the universe of features that operate in other cultures. We share that concern and accept the value of qualitative, in-depth studies of a single culture.

However, the aim of this book is not to understand how any one particular culture functions. Because ethnic groups differ appreciably with respect to both their own characteristics and the patterns of psychological and social functioning with which they are associated, we sought to use variations across ethnic groups as a means of examining and, where possible, testing competing hypotheses about the mediating causal mechanisms. Accordingly, to somewhat oversimplify, the research strategy involves determining which risk or protective factors are significantly associated with particular psychosocial outcomes within each ethnic group. We reasoned that if specific factors mediate an observed between-group difference, then the prevalence of the risk/protective factor will differ between the ethnic groups; moreover, when the factor is introduced into a multivariate analysis, it will obliterate (or, at least, bring about a major reduction in) the difference between groups.