Causal discovery seeks information about causal structure from observations or interventions, rather than assuming the complete structure in advance. In a causal directed acyclic graph, observed conditional independences can restrict possible arrows, but generally do not identify a unique graph without further assumptions.
For three variables, the chains and , and the fork , all imply and are independent conditional on in the corresponding faithful models. These independences alone cannot distinguish those structures. By contrast, the collider generically makes and marginally independent but dependent after conditioning on .
This example explains both the appeal and the limitation of searching for graphical structure. Conditional independence tests offer clues, while hidden common causes, selection and unfaithful parameter cancellations can undermine the inference. An experiment that changes can distinguish directions that observational data alone leave ambiguous.[ref]
Nested Markov Properties for Acyclic Directed Mixed Graphs by Thomas S. Richardson and others. Introduction, hidden-variable models and observable constraints
New to topics? Read the docs here!