TIES THAT BIND: Measurement, Demographics, and Social Connectedness*

    loading  Checking for direct PDF access through Ovid


In this article I develop a taxonomy of how demographic variables are used in political science models. The functions of demographics—for description, as controls in statistical models, or as proxies for external societal cleavages or underlying individual attributes—raise questions about validity and responsible usage. To illustrate the more general problem, the construct of social connectedness is examined in regard to its relationship with various demographic variables and its impact on voter turnout. Using data from the 1992 National Election Study, the analyses indicate that marital status, church attendance, owning a home, formal group membership, education, and income are all related to social connectedness, but the impact of education and income on electoral participation does not appear to be mediated by social connectedness. The results illustrate that careful consideration of measurement issues can clarify the relationships in our substantive models and that substantive models can illuminate measurement issues as well.

Related Topics

    loading  Loading Related Articles