The Annals of Statistics

Markov equivalence for ancestral graphs

R. Ayesha Ali, Thomas S. Richardson, and Peter Spirtes

Full-text: Open access


Ancestral graphs can encode conditional independence relations that arise in directed acyclic graph (DAG) models with latent and selection variables. However, for any ancestral graph, there may be several other graphs to which it is Markov equivalent. We state and prove conditions under which two maximal ancestral graphs are Markov equivalent to each other, thereby extending analogous results for DAGs given by other authors. These conditions lead to an algorithm for determining Markov equivalence that runs in time that is polynomial in the number of vertices in the graph.

Article information

Ann. Statist. Volume 37, Number 5B (2009), 2808-2837.

First available in Project Euclid: 17 July 2009

Permanent link to this document

Digital Object Identifier

Mathematical Reviews number (MathSciNet)

Zentralblatt MATH identifier

Primary: 68T30: Knowledge representation 05C75: Structural characterization of families of graphs
Secondary: 68T37: Reasoning under uncertainty

Directed acyclic graphs discriminating path inducing path Markov equivalence polynomial-time algorithm


Ali, R. Ayesha; Richardson, Thomas S.; Spirtes, Peter. Markov equivalence for ancestral graphs. Ann. Statist. 37 (2009), no. 5B, 2808--2837. doi:10.1214/08-AOS626.

Export citation


  • [1] Ali, R. A., Richardson, T. and Spirtes, P. (2009). Supplementary material for “Markov equivalence for ancestral graphs.” Available at
  • [2] Andersson, S. A., Madigan, D. and Perlman, M. D. (1997). A characterization of Markov equivalence classes for acyclic digraphs. Ann. Statist. 25 505–541.
  • [3] Chickering, D. (1995). A transformational characterization of equivalent Bayesian network. In Uncertainty in Artificial Intelligence: Proceedings of the 11th Conference (P. Besnard and S. Hanks, eds.) 87–98. Morgan Kaufmann, San Francisco.
  • [4] Cooper, G. F. (1995). Causal discovery from data in the presence of selection bias. In Preliminary Papers, Fifth International Workshop on AI and Statistics (D. Fisher, ed.) 140–150. Ft. Lauderdale, FL.
  • [5] Cox, D. R. and Wermuth, N. (1996). Multivariate Dependencies: Models, Analysis and Interpretation. Chapman and Hall, London.
  • [6] Dawid, A. (1980). Conditional independence for statistical operations. Ann. Statist. 8 598–617.
  • [7] Drton, M., Andersson, S. A. and Perlman, M. D. (2006). Conditional independence models for seemingly unrelated regressions with incomplete data. J. Multivariate Anal. 97 385–411.
  • [8] Drton, M. and Richardson, T. S. (2004). Multimodality of the likelihood in the bivariate seemingly unrelated regressions model. Biometrika 91 383–392.
  • [9] Frydenberg, M. (1990). The chain graph Markov property. Scand. J. Statist. 17 333–353.
  • [10] Koster, J. T. A. (1999). Linear Structural Equations and Graphical Models. The Fields Institute, Toronto.
  • [11] Koster, J. T. A. (2002). Marginalizing and conditioning in graphical models. Bernoulli 8 817–840.
  • [12] Lauritzen, S. (1996). Graphical Models. Oxford Statistical Science Series 17. Clarendon, Oxford, UK.
  • [13] Meek, C. (1995). Causal inference and causal explanation with background knowledge. In Uncertainty in Artificial Intelligence: Proceedings of the 11th Conference (P. Besnard and S. Hanks, eds.) 403–410. Morgan Kaufmann, San Francisco.
  • [14] Pearl, J. (2000). Causality: Models, Reasoning and Inference. Cambridge Univ. Press, Cambridge, UK.
  • [15] Richardson, T. S. and Spirtes, P. (2002). Ancestral graph Markov models. Ann. Statist. 30 962–1030.
  • [16] Richardson, T. S. and Spirtes, P. (2003). Causal inference via ancestral graph models. In Highly Structured Stochastic Systems (P. Green, N. Hjort and S. Richardson, eds.). Oxford Univ. Press, Oxford.
  • [17] Robins, J. (1997). Causal inference from complex longitudinal data. In Latent Variable Modeling and Applications to Causality (Los Angeles, CA, 1994). Lecture Notes in Statistics 120 (M. Berkane, ed.) 69–117. Springer, New York.
  • [18] Spirtes, P., Glymour, C. and Scheines, R. (1993). Causation, Prediction and Search. Lecture Notes in Statistics 81. Springer.
  • [19] Spirtes, P. and Richardson, T. S. (1997). A polynomial-time algorithm for determining DAG equivalence in the presence of latent variables and selection bias. In Preliminary Papers, Sixth International Workshop on AI and Statistics (D. Madigan and P. Smyth, eds.) 489–501. Ft. Lauderdale, FL.
  • [20] Studeny, M. (2004). Probabilistic Conditional Independence Structures. Springer, Santa Clara, CA, USA.
  • [21] Verma, T. and Pearl, J. (1990). Equivalence and synthesis of causal models. In Uncertainty in Artificial Intelligence: Proceedings of the Sixth Conference (M. Henrion, R. Shachter, L. Kanal and J. Lemmer, eds.) 220–227. Association for Uncertainty in AI, Mountain View, CA.
  • [22] Verma, T. and Pearl, J. (1991). Equivalence and synthesis of causal models. Technical Report R-150, Cognitive Systems Laboratory, UCLA.
  • [23] Zellner, A. (1962). An efficient method of estimating seemingly unrelated regression equations and tests for aggregation bias. J. Amer. Statist. Assoc. 57 348–368.
  • [24] Zhao, H., Zheng, Z. and Liu, B. (2005). On the Markov equivalence of maximal ancestral graphs. Sci. China Ser. A 48 548–562.