Humboldt-Universität zu Berlin - Department of Social Sciences - Berlin Graduate School of Social Science

Alejandra Rodriguez

Name
Alejandra Rodriguez
Institut für Sozialwissensachften, Humboldt Universität zu Berlin Cohort

BGSS Generation 2017

 

Title

"Family Behavior and Children's Outcomes: Modeling and Measurement Issues"

 

Supervisor

Prof. Dr. Anette Fasang

 

Abstract

In the study of social inequality, family behavior has gained salience as a potential cause in the intergenerational transmission of inequality. For example, changes in family formation and dissolution processes, parenting, and single events such as divorce have been considered as potential causes that perpetuate inequality across generations. However, given that family behavior is an endogenous variable and that studies of these topics rely on observational data, causal inference in family research must account for the multiple ways in which endogeneity may bias causal estimates. This dissertation focuses on a set of four methodological challenges that generate endogeneity bias: (1) family instability as a time-varying exposure affected by “treatment-confounder feedback bias”; (2) divorce or separation effects as affected by life course selection bias; (3) parenting as a causal mediation process; and (4) measurement error as a pervasive form of endogeneity when studying unobservable children’s cognitive and noncognitive outcomes. The solutions proposed in the dissertation involve the use of marginal structural models, machine learning algorithms that improve causal inference, causal mediation methods that account for interactions among mediating mechanisms, and a framework for exploring forms of measurement error in unobservable constructs such as children’s cognitive abilities. In addition to that, a final chapter of the dissertation reflects on family behavior as a normative topic that influences how previous research has thought (or not thought) about it, and on how this, in turn, mirrors in the analyses and the policy conclusions that are drawn by researchers.  Estimation of causal effects remains a challenging task for family demography scholars, but this dissertation proposes and explores four ways in which such challenges can be individually addressed. The literature on family inequality should reconsider the extent to which family behavior is not much of a cause of negative effects and more of an effect of persistent social inequalities.

 

Recent Publications
  • Rodríguez Sánchez, A. (2020). Fair comparisons: life course selection bias and the effect of father absence on US children. Working paper (under review)
  • Rodríguez Sánchez, A. (2020). What can parents do? SES and heterogeneous parenting effects on children's language. Working paper (under review)
  • Rodríguez Sánchez, A. (2020). Is it just noise? Measuring unobservable cognitive abilities in early childhood. Personality and Individual Differences, 166, 110162.
  • Rodríguez Sánchez, A. (2019). Family structure effects on US children’s well-being? Re-examining the family instability hypothesis (Working paper), No. 84q56, Center for Open Science.

 

Institution