Path Dependent Structural Equation Models

Abstract

Causal analyses of longitudinal data generally assume that the qualitative causal structure relating variables remains invariant over time. In structured systems that transition between qualitatively different states in discrete time steps, such an approach is deficient on two fronts. First, time-varying variables may have state-specific causal relationships that need to be captured. Second, an intervention can result in state transitions downstream of the intervention different from those actually observed in the data. In other words, interventions may counterfactually alter the subsequent temporal evolution of the system. We introduce a generalization of causal graphical models, Path Dependent Structural Equation Models (PDSEMs), that can describe such systems. We show how causal inference may be performed in such models and illustrate its use in simulations and data obtained from a septoplasty surgical procedure.

References in corpus (5)

• Likelihood ratio tests and singularities
• Nested Markov Properties for Acyclic Directed Mixed Graphs
• Markovian acyclic directed mixed graphs for discrete data
• Computing maximum likelihood estimates in recursive linear models with correlated errors
• Semiparametric Inference For Causal Effects In Graphical Models With Hidden Variables