Glossary · Marketing Engineering

Directed Acyclic Graph (Causal DAG)

also: DAG · Bayesian network · structural causal model

Definition

A Directed Acyclic Graph is the formal representation of a causal structure: nodes are variables, directed edges are direct causal effects, and no cycles are permitted. DAGs encode the identification assumptions needed to estimate causal effects from observational data via the back-door and front-door criteria.

Pearl's causal DAG formalism provides the grammar of causal inference. Three node roles: chains (mediators), forks (common causes / confounders), and colliders (common effects). The back-door criterion identifies sufficient adjustment sets to block confounding paths; the front-door criterion identifies causal effects through mediators when direct confounding is unobserved. In marketing attribution, DAGs replace ad-hoc rules with a transparent assumption layer.

Essays on this concept