From statistical associations to causation: What developmentalists can learn from instrumental variables techniques coupled with experimental data.

Abstract

In this article, the authors aim to make accessible the careful application of a method called instrumental variables (IV). Under the right analytic conditions, IV is one promising strategy for answering questions about the causal nature of associations and, in so doing, can advance developmental theory. The authors build on prior work combining the analytic approach of IV with the strengths of random assignment design, whether the experiment is conducted in the lab setting or in the "real world." The approach is detailed through an empirical example about the effects of maternal education on children's cognitive and school outcomes. With IV techniques, the authors address whether maternal education is causally related to children's cognitive development or whether the observed associations reflect some other characteristic related to parenting, income, or personality. The IV estimates show that maternal education has a positive effect on the cognitive test scores of children entering school. The authors conclude by discussing opportunities for applying these same techniques to address other questions of critical relevance to developmental science.

What makes articles highly cited?

“…Any observed relationship may be partly or fully driven by omitted variables and the estimated relation could be higher, lower, or of a different sign than the actual causal relation. The confounding effect of endogeneity on estimates is one of the biggest methodological problems in management and psychology research today (Antonakis, et al, 2010;Bascle, 2008;Billings & Wroten, 1978;Duncan, Magnusson, & Ludwig, 2004;Gennetian, Magnuson, & Morris, 2008;Hamilton & Nickerson, 2003;Larcker & Rusticus, 2010). It is a problem that has mostly been ignored by researchers in the past and mostly plagues non-experimental research (but also experimental research too, for example, when testing mediation in the case of an endogenous mediator).…”

“…Researchers must correctly model this source of endogeneity bias by using two-stage least squares regression or by allowing the disturbances of the endogenous variables to correlate in the context of a SEM model (refer to the following for detailed explanations on testing mediation correctly: Antonakis, et al, 2010;Foster, 2010;Foster & McLanahan, 1996;Gennetian, et al, 2008;Shaver, 2005).…”

Preschool center care quality effects on academic achievement: An instrumental variables analysis.

“…However, several studies have demonstrated the effectiveness of analyzing multisite experimental data by using treatment and treatment-by-site interactions as instruments in an IV analysis (Bloom, Zhu, & Unlu, 2010;Duncan, Morris, & Rodrigues, 2011;Gennetian et al, 2008;Ludwig & Kling, 2006), and recent studies have applied this methodology to the study of the determinants of children's development (Crosby et al, 2010;Duncan et al, 2011;Gennetian et al, 2008).…”

“…This approach is predominately used in economics (e.g., Angrist & Krueger, 1991Angrist & Pischke, 2008); however, there is increased interest in using this statistical technique to test for causal effects within developmental research (e.g., Crosby et al, 2010;Gennetian et al, 2008). The IV approach requires the identification of variables (instruments) that are moderately to strongly related to the predictor of interest and that serve as the only conduit through which the predictor has an impact on the outcome.…”

“…None of the previous studies, however, have been able to account for all observed and unobserved potential sources of bias and thus have not been able to confidently estimate causal effects. The purpose of this study was to demonstrate how an econometric technique, instrumental variables (IV), can be used with multisite experimental data to estimate causal effects for topics within child development that cannot be tested in random assignment studies (Crosby, Dowsett, Gennetian, & Huston, 2010;Gennetian, Magnuson, & Morris, 2008). We used the example of understanding the effect of center care quality on preschool children's academic achievement.…”