Thinking Clearly About Correlations and Causation: Graphical Causal Models for Observational Data

Julia M Rohrer

doi:10.1177/2515245917745629

Abstract

Correlation does not imply causation; but often, observational data are the only option, even though the research question at hand involves causality. This article discusses causal inference based on observational data, introducing readers to graphical causal models that can provide a powerful tool for thinking more clearly about the interrelations between variables. Topics covered include the rationale behind the statistical control of third variables, common procedures for statistical control, and what can go wrong during their implementation. Certain types of third variables—colliders and mediators—should not be controlled for because that can actually move the estimate of an association away from the value of the causal effect of interest. More subtle variations of such harmful control include using unrepresentative samples, which can undermine the validity of causal conclusions, and statistically controlling for mediators. Drawing valid causal inferences on the basis of observational data is not a mechanistic procedure but rather always depends on assumptions that require domain knowledge and that can be more or less plausible. However, this caveat holds not only for research based on observational data, but for all empirical research endeavors.

Highlights

Psychologists in many fields face a dilemma
Universitat Leipzig Fakultat fur Biowissenschaften Pharmazie und Psychologie–Psychology, Neumarkt 9-19, Leipzig 04103, Germany E-mail: julia.rohrer@uni-leipzig.de. Rohrer assessment of their social class (Piff, Kraus, Côté, Cheng, & Keltner, 2010). Using such surrogates can result in valuable insights, but they are not a panacea as they come with a well-known trade-off (e.g., Cook & Campbell, 1979): they substantially improve confidence in the internal validity of a study, they might substantially decrease the external validity; that is, it becomes uncertain whether the finding says much about other situations, other operationalizations of the independent variable, or the world outside the lab in general
The practice of making causal inferences on the basis of observational data depends crucially on awareness of potential confounders and meaningful statistical control that takes into account estimation issues such as nonlinear confounding and measurement error

Summary

A Brief Introduction to Directed Acyclic Graphs

Assume that we are interested in the causal effect of educational attainment on income. In this sense, a DAG is qualitative: A → B means only that A causally affects B in some way. From these two simple building blocks—nodes and arrows—one can visualize more complex situations and trace paths from variable to variable Intelligence and income are connected by the paths intelligence → educational attainment

30 Grades

Conclusion

Declaration of Conflicting Interests