• Bernoulli
  • Volume 17, Number 3 (2011), 845-879.

Probability distributions with summary graph structure

Nanny Wermuth

Full-text: Open access


A set of independence statements may define the independence structure of interest in a family of joint probability distributions. This structure is often captured by a graph that consists of nodes representing the random variables and of edges that couple node pairs. One important class contains regression graphs. Regression graphs are a type of so-called chain graph and describe stepwise processes, in which at each step single or joint responses are generated given the relevant explanatory variables in their past. For joint densities that result after possible marginalising or conditioning, we introduce summary graphs. These graphs reflect the independence structure implied by the generating process for the reduced set of variables and they preserve the implied independences after additional marginalising and conditioning. They can identify generating dependences that remain unchanged and alert to possibly severe distortions due to direct and indirect confounding. Operators for matrix representations of graphs are used to derive these properties of summary graphs and to translate them into special types of paths in graphs.

Article information

Bernoulli, Volume 17, Number 3 (2011), 845-879.

First available in Project Euclid: 7 July 2011

Permanent link to this document

Digital Object Identifier

Mathematical Reviews number (MathSciNet)

Zentralblatt MATH identifier

concentration graph directed acyclic graph endogenous variables graphical Markov model independence graph multivariate regression chain partial closure partial inversion triangular system


Wermuth, Nanny. Probability distributions with summary graph structure. Bernoulli 17 (2011), no. 3, 845--879. doi:10.3150/10-BEJ309.

Export citation


  • Andersson, S.A., Madigan, D., Perlman, M.D. and Triggs, C.M. (1997). A graphical characterization of lattice conditional independence models. Ann. Math. Artif. Intell. 21 27–50.
  • Barndorff-Nielsen, O.E. (1978). Information and Exponential Families in Statistical Theory. Chichester: Wiley.
  • Brito, C. and Pearl, J. (2002). A new identification condition for recursive models with correlated errors. Struct. Equ. Model. 9 459–474.
  • Brown, L.D. (1986). Fundamentals of Statistical Exponential Families with Applications in Statistical Decision Theory. LNMS 9. Hayward, CA: Inst. Math. Statist.
  • Castelo, R. and Siebes, A. (2003). A characterization of moral transitive acyclic directed graph Markov models as labeled trees. J. Statist. Plann. Inference 115 235–259.
  • Cox, D.R. (2006). Principles of Statistical Inference. Cambridge: Cambridge Univ. Press.
  • Cox, D.R. and Wermuth, N. (1993). Linear dependencies represented by chain graphs (with discussion). Statist. Sci. 8 204–218, 247–277.
  • Cox, D.R. and Wermuth, N. (1996). Multivariate Dependencies: Models, Analysis, and Interpretation. London: Chapman and Hall.
  • Cox, D.R. and Wermuth, N. (2000). On the generation of the chordless 4-cycle. Biometrika 87 206–212.
  • Cox, D.R. and Wermuth, N. (2004). Causality: A statistical view. Internat. Statist. Rev. 72 285–305.
  • Dawid, A.P. (1979). Some misleading arguments involving conditional independence. J. Roy. Statist. Soc. B 41 249–252.
  • Dempster, A.P. (1972). Covariance selection. Biometrics 28 157–175.
  • Dirac, G.A. (1961). On rigid circuit graphs. Abhandl. Math. Sem. Hamburg 25 71–76.
  • Drton, M. (2009). Discrete chain graph models. Bernoulli 15 736–753.
  • Drton, M., Eichler, M. and Richardson, T.S. (2009). Computing maximum likelihood estimates in recursive linear models. J. Mach. Learn. Res. 10 2329–2348.
  • Frydenberg, M. and Lauritzen, S.L. (1989). Decomposition of maximum likelihood in mixed interaction models. Biometrika 76 539–555.
  • Geiger, D., Verma, T.S. and Pearl, J. (1990). Identifying independence in Bayesian networks. Networks 20 507–534.
  • Holland, P.W. and Rosenbaum, P.R. (1989). Conditional association and unidimensionality in monotone latent variable models. Ann. Statist. 14 1523–1543.
  • Kang, C. and Tian, J. (2009). Markov properties for linear causal models with correlated errors. J. Mach. Learn. Res. 10 41–70.
  • Kauermann, G. (1996). On a dualization of graphical Gaussian models. Scand. J. Statist. 23 115–116.
  • Koster, J.T.A. (2002). Marginalising and conditioning in graphical models. Bernoulli 8 817–840.
  • Lauritzen, S.L. (1996). Graphical Models. Oxford: Oxford Univ. Press.
  • Lauritzen, S.L. and Wermuth, N. (1989). Graphical models for association between variables, some of which are qualitative and some quantitative. Ann. Statist. 17 31–57.
  • Lehmann, E.L. and Scheffé, H. (1955). Completeness, similar regions and unbiased estimation. Sankhyā A 14 219–236.
  • Mandelbaum, A. and Rüschendorf, L. (1987). Complete and symmetrically complete families of distributions. Ann. Statist. 15 1229–1244.
  • Marchetti, G.M. and Lupparelli, M. (2011). Chain graph models of multivariate regression type for categorical data. To appear.
  • Marchetti, G.M. and Wermuth, N. (2009). Matrix representations and independencies in directed acyclic graphs. Ann. Statist. 47 961–978.
  • Neumann, C.G. (1884). Vorlesungen über Riemann’sche Theorie der Abel’schen Integrale, 2nd ed. Leipzig: Teubner.
  • Paz, A. (2007). The annotated graph model for representing DAG-representable relations – algorithms approach. Research report, Technion, Haifa.
  • Pearl, J. (1988). Probabilistic Reasoning in Intelligent Systems. San Mateo: Morgan and Kaufmann.
  • Pearl, J. and Wermuth, N. (1994). When can association graphs admit a causal interpretation? In Models and Data, Artificial Intelligence and Statistics IV (P. Cheeseman and W. Oldford, eds.) 205–214. New York: Springer.
  • Richardson, T.S. and Spirtes, P. (2002). Markov graph ancestral models. Ann. Statist. 30 962–1030.
  • Robins, J. and Wasserman, L. (1997). Estimation of effects of sequential treatments by reparametrizing directed acyclic graphs. In Proceedings, 13th Annual Conference on Uncertainty in Artificial Intelligence (D. Geiger and O. Shenoy, eds.) 409–420. San Francisco: Morgan and Kaufmann.
  • Sadeghi, K. (2009). Representing modified independence structures. Transfer thesis, Oxford Univ.
  • San Martin, E. and Mochart, M. (2007). On joint completeness: Sampling and Bayesian versions, and their connections. Sankhyā A 69 780–807.
  • San Martin, E., Mochart, M. and Rolin, J.M. (2005). Ignorable common information, null sets and Basu’s first theorem. Sankhyā A 67 674–698.
  • Sargan, J.D. (1958). The estimation of economic relationships using instrumental variables. Econometrica 26 393–415.
  • Speed, T.P. and Kiiveri, H.T. (1986). Gaussian Markov distributions over finite graphs. Ann. Statist. 14 138–150.
  • Spirtes, P., Glymour, C. and Scheines, R. (1993). Causation, Prediction and Search. New York: Springer.
  • Studený, M. (2005). Probabilistic Conditional Independence Structures. London: Springer.
  • Tikhonov, A.N. (1963). Solution of ill-posed problems and the regularization method (in Russian). Dokl. Akad. Nauk SSSR 153 49–52.
  • Wermuth, N. and Cox, D.R. (1998). On association models defined over independence graphs. Bernoulli 4 477–495.
  • Wermuth, N. and Cox, D.R. (2004). Joint response graphs and separation induced by triangular systems. J. R. Stat. Soc. Ser. B Stat. Methodol. 66 687–717.
  • Wermuth, N. and Cox, D.R. (2008). Distortions of effects caused by indirect confounding. Biometrika 95 17–33.
  • Wermuth, N., Cox, D.R. and Pearl, J. (1994). Explanantions for multivariate structures derived from univariate recursive regressions. In Ber. Stoch. Verw. Geb., Univ. Mainz 94. ISSN-0098, Mainz.
  • Wermuth, N. and Lauritzen, S.L. (1990). On substantive research hypotheses, conditional independence graphs and graphical chain models (with discussion). J. Roy. Statist. Soc. Ser. B 52 21–75.
  • Wermuth, N., Marchetti, G.M. and Cox, D.R. (2009). Triangular systems for symmetric binary variables. Electron. J. Statist. 3 932–955.
  • Wermuth, N. and Sadeghi, K. (2011). Sequences of regressions and their independences. To appear.
  • Wermuth, N., Wiedenbeck, M. and Cox, D.R. (2006). Partial inversion for linear systems and partial closure of independence graphs. Numer. Math. 46 883–901.
  • Wiedenbeck, M. and Wermuth, N. (2010). Changing parameters by partial mappings. Statist. Sinica 20 932–955.