Utilizing causal diagrams across quasi-experimental approaches

Abstract

Recent developments in computer science have substantially advanced the use of observational causal inference under Pearl's structural causal model (SCM) framework. A key tool in the application of SCM is the use of casual diagrams, used to visualize the causal structure of a system or process under study. Here, we show how causal diagrams can be extended to ensure proper study design under quasi-experimental settings, including propensity score analysis, before-after-control-impact studies, regression discontinuity design, and instrumental variables. Causal diagrams represent a unified approach to variable selection across methodologies and should be routinely applied in ecology research with causal implications.