A Mixed-Methods Social Networks Study Design for Research on Transnational Families

Article excerpt

This paper advocates the adoption of a mixed-methods research design to describe and analyze ego-centered social networks in transnational family research. Drawing on the experience of the "Social Networks Influences on Family Formation" project (2004-2005; see Bernardi, Keim, & von der Lippe, 2007a, 2007b), I show how the combined use of network generators and semistructured interviews (N = 116) produces unique data on family configurations and their impact on life course choices. A mixed-methods network approach presents specific advantages for research on children in transnational families. On one hand, quantitative analyses are crucial for reconstructing and measuring the potential and actual relational support available to children in a context where kin interactions may be hindered by temporary and prolonged periods of separation. On the other hand, qualitative analyses can address strategies and practices employed by families to maintain relationships across international borders and geographic distance, as well as the implications of those strategies for children's well-being.

Key Words: children, family relations, fertility, immigrant families, social networks, transilion to parenthood.

An empirical gap exists in knowledge concerning the effect of transnational family life on the well-being of children. Although family sociology discusses the implications of family ties other than those among coresident members, surveys commonly treat migrant family ties as though the families live together, and the data are collected accordingly. Family researchers and policy makers, however, need to take into account the spatial distribution of families and the effects of distance between family members. International migration affects the lives of both family members who migrate and those who remain behind, having important consequences for kinship ties, living arrangements, and children's well-being.

Transnational families are, by definition, spread across geographical and legal borders, whereas transnationalism has been conceptualized as "a set of sustained long-distance, border-crossing connections" (Vertovec, 2004, p. 3) or "practices and relationships that link migrants and their children with the home country, where such practices have significant meaning and are regularly observed" (Levitt & Jaworsky, 2007, p. 132). Transnational families, nuclear or extended, are dispersed across international borders, comprising family members who spend time in one country or another, depending on a variety of factors such as work, education, legal requirements for residence permits, and care and support for other family members. These families often depend on a cross-border division of labor in which production occurs in the host country and reproduction occurs in the country of origin (Schmalzbauer, 2004). The families are simultaneously incorporated into the host and origin societies and negotiate their identities by maintaining multiple connections. In this sense, the definition of transnational applies to families and individuals alike when they are embedded in an interdependent set of family relationships spread across different nations. A typical case, having arisen with the feminization of migration flows, is the situation involving women who do their mothering at a distance and send their wages home (Ehrenreich & Hochschild, 2002). Their families and often their extended families depend on them for sustenance. At the same time, these women depend on their social networks to care for and raise their children for short or long periods. The actual practice of mothering is often assumed by aunts, grandmothers, other kin, and neighbors (Aranda, 2003; Moon, 2003; Salazar Parreñas, 2001).

As Mazzucato and Schans (2011) have observed, studies concerning the manner in which children's care and socialization are organized in transnational families are just beginning to go beyond the analysis of mother- children and parent - children bonds and qualifying as relational subsystems (Cowan, Cowan, Herring, & Miller, 1991). …