Causal Inference in the Health Sciences: A Conceptual Introduction

  • Judea Pearl


This paper provides a conceptual introduction to causal inference, aimed to assist health services researchers benefit from recent advances in this area. The paper stresses the paradigmatic shifts that must be undertaken in moving from traditional statistical analysis to causal analysis of multivariate data. Special emphasis is placed on the assumptions that underlie all causal inferences, the languages used in formulating those assumptions, and the conditional nature of causal claims inferred from nonexperimental studies. These emphases are illustrated through a brief survey of recent results, including the control of confounding, corrections for noncompliance, and a symbiosis between counterfactual and graphical methods of analysis.

structural equation models confounding noncompliance graphical methods counterfactuals 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. J. D. Angrist, G. W. Imbens and D. B. Rubin, “Identification of causal effects using instrumental variables (with comments),” Journal of the American Statistical Association, 91(434), pp. 444-472, 1996.Google Scholar
  2. A. Balke and J. Pearl, “Counterfactuals and policy analysis in structural models,” in Uncertainty in Artificial Intelligence 11 (P. Besnard and S. Hanks, eds.), Morgan Kaufmann, San Francisco, pp. 11-18, 1995.Google Scholar
  3. A. Balke and J. Pearl, “Bounds on treatment effects from studies with imperfect compliance,” Journal of the American Statistical Association, 92(439), pp. 1172-1176, 1997.Google Scholar
  4. H. Becher, “The concept of residual confounding in regression models and some applications,” Statistics in Medicine, 11, pp. 1747-1758, 1992.Google Scholar
  5. Y. M. M. Bishop, “Effects of collapsing multidimensional contingency tables,” Biometrics, 27, pp. 545-562, 1971.Google Scholar
  6. K. A. Bollen. Structural Equations with Latent Variables, John Wiley, New York, 1989.Google Scholar
  7. B. Bonet, “Instrumentality tests revisited,” in Proceedings of the Seventeenth Conference on Uncertainty in Artificial Intelligence, Morgan Kaufmann, San Francisco, CA, pp. 48-55, 2001.Google Scholar
  8. R. J. Bowden and D. A. Turkington. Instrumental Variables, Cambridge University Press, Cambridge, England, 1984.Google Scholar
  9. N. E. Breslow and N. E. Day. Statistical Methods in Cancer Research; Vol. 1, The Analysis of Case-Control Studies, IARC, Lyon, 1980.Google Scholar
  10. N. Cartwright. Nature's Capacities and Their Measurement, Clarendon Press, Oxford, 1989.Google Scholar
  11. D. M. Chickering and J. Pearl, “A clinician's tool for analyzing non-compliance,” Computing Science and Statistics, 29(2), pp. 424-431, 1997.Google Scholar
  12. R. G. Cowell, A. P. Dawid, S. L. Lauritzen and D. J. Spielgelhalter. Probabilistic Networks and Expert Systems, Springer Verlag, New York, NY, 1999.Google Scholar
  13. D. R. Cox. The Planning of Experiments, John Wiley and Sons, NY, 1958.Google Scholar
  14. A. P. Dawid, “Conditional independence in statistical theory,” Journal of the Royal Statistical Society, Series B, 41(1), pp. 1-31, 1979.Google Scholar
  15. O. D. Duncan. Introduction to Structural Equation Models, Academic Press, New York, 1975.Google Scholar
  16. E. Eells. Probabilistic Causality, Cambridge University Press, Cambridge, MA, 1991.Google Scholar
  17. D. Freedman, “As others see us: A case study in path analysis (with discussion),” Journal of Educational Statistics, 12(2), pp. 101-223, 1987.Google Scholar
  18. D. Galles and J. Pearl, “Testing identifiability of causal effects,” in Uncertainty in Artificial Intelligence 11 (P. Besnard and S. Hanks, eds.), Morgan Kaufmann, San Francisco, pp. 185-195, 1995.Google Scholar
  19. A. S. Goldberger, “Structural equation models in the social sciences,” Econometrica: Journal of the Econometric Society, 40, pp. 979-1001, 1972.Google Scholar
  20. D. A. Grayson, “Confounding confounding,” American Journal of Epidemiology, 126, pp. 546-553, 1987.Google Scholar
  21. S. Greenland and J. M. Robins, “Identifiability, exchangeability, and epidemiological confounding,” International Journal of Epidemiology, 15(3), pp. 413-419, 1986.Google Scholar
  22. S. Greenland, J. Pearl and J. M Robins, “Causal diagrams for epidemiologic research,” Epidemiology, 10(1), pp. 37-48, 1999a.Google Scholar
  23. S. Greenland, J. M. Robins and J. Pearl, “Confounding and collapsibility in causal inference,” Statistical Science, 14(1), pp. 29-46, 1999b.Google Scholar
  24. S. Greenland, “Relation of the probability of causation to the relative risk and the doubling dose: A methodologic error that has become a social problem. American Journal of Public Health, 89, pp. 1166-1169, 1999.Google Scholar
  25. W. W. Hauck, J. M. Heuhaus, J. D. Kalbfleisch and S. Anderson, “A consequence of omitted covariates when estimating odds ratios,” Journal Clinical Epidemiology, 44(1), pp. 77-81, 1991.Google Scholar
  26. J. J. Heckman and J. Smith, “Evaluating the welfare state,” in Econometric and Economic Theory in the 20th Century (S. Strom, ed.), Cambridge University Press, Cambridge, England, pp. 1-60, 1998.Google Scholar
  27. P. W. Holland and D. B. Rubin, “Causal inference in retrospective studies,” Evaluation Review, 13, pp. 203-231, 1988.Google Scholar
  28. P. W. Holland, “Causal inference, path analysis, and recursive structural equations models,” in Sociological Methodology (C. Clogg, ed.), American Sociological Association, Washington, D.C., pp. 449-484, 1988.Google Scholar
  29. G. W. Imbens and J. D. Angrist, “Identification and estimation of local average treatment effects,” Econometrica, 62(2), pp. 467-475, 1994.Google Scholar
  30. K. G. Joreskog and D. Sorbom. LISREL IV: Analysis of Linear Structural Relationships by Maximum Likelihood, International Educational Services, Chicago, 1978.Google Scholar
  31. J. S. Kaufman and S. Kaufman, “Assessment of structured socioeconomic effects on health,” Epidemiology, 12(2), pp. 157-167, 2001.Google Scholar
  32. H. Kiiveri, T. P. Speed and J. B. Carlin, “Recursive causal models,” Journal of Australian Math Society, 36, pp. 30-52, 1984.Google Scholar
  33. D. G. Kleinbaum, L. L. Kupper, K. E. Muller and A. Nizam. Applied Regression Analysis and Other Multivariable Methods, Duxbury Press, Pacific Grove, third edition, 1998.Google Scholar
  34. T. C. Koopmans, “Identification problems in econometric model construction,” in Studies in Econometric Method (W. C. Hood and T. C. Koopmans, eds.), Wiley, New York, pp. 27-48, 1953.Google Scholar
  35. M. Kuroki and M. Miyakawa, “Identifiability criteria for causal effects of joint interventions,” Journal of the Japan Statistical Society, 29(2), pp. 105-117, 1999.Google Scholar
  36. S. L. Lauritzen. Graphical Models, Clarendon Press, Oxford, 1996.Google Scholar
  37. S. L. Lauritzen, “Causal inference from graphical models,” Technical Report R-99-2021, Department of Mathematical Sciences, Aalborg University, Denmark, 1999.Google Scholar
  38. D. V. Lindley and M. R. Novick, “The role of exchangeability in inference,” The Annals of Statistics, 9(1), pp. 45-58, 1981.Google Scholar
  39. C. F. Manski, “Nonparametric bounds on treatment effects,” American Economic Review, Papers and Proceedings, 80, pp. 319-323, 1990.Google Scholar
  40. C. F. Manski. Identification Problems in the Social Sciences, Harvard University Press, Cambridge, MA, 1995.Google Scholar
  41. O. S. Miettinen and E. F. Cook, “Confounding essence and detection,” American Journal of Epidemiology, 114, pp. 593-603, 1981.Google Scholar
  42. B. Muthen, “Response to Freedman's critique of path analysis: Improve credibility by better methodological training,” Journal of Educational Statistics, 12(2), pp. 178-184, 1987.Google Scholar
  43. J. Neyman, “On the application of probability theory to agricultural experiments,” Essay on principles. Section 9. Statistical Science, 5(4), pp. 465-480, 1923.Google Scholar
  44. J. Pearl and J. M. Robins, “Probabilistic evaluation of sequential plans from causal models with hidden variables,” in Uncertainty in Artificial Intelligence 11 (P. Besnard and S. Hanks, eds.), Morgan Kaufmann, San Francisco, pp. 444-453, 1995.Google Scholar
  45. J. Pearl and T. Verma, “A theory of inferred causation,” in Principles of Knowledge Representation and Reasoning: Proceedings of the Second International Conference (J. A. Allen, R. Fikes and E. Sandewall, eds.), Morgan Kaufmann, San Mateo, CA, pp. 441-452, 1991.Google Scholar
  46. J. Pearl. Probabilistic Reasoning in Intelligent Systems, Morgan Kaufmann, San Mateo, CA, 1988.Google Scholar
  47. J. Pearl, “Comment: Graphical models, causality, and intervention,” Statistical Science, 8, pp. 266-269, 1993.Google Scholar
  48. J. Pearl, “Causal diagrams for empirical research,” Biometrika, 82(4), pp. 669-710, 1995a.Google Scholar
  49. J. Pearl, “On the testability of causal models with latent and instrumental variables,” in Uncertainty in Artificial Intelligence 11 (P. Besnard and S. Hanks, eds.), Morgan Kaufmann, pp. 435-443, 1995b.Google Scholar
  50. J. Pearl. Causality: Models, Reasoning, and Inference, Cambridge University Press, New York, 2000.Google Scholar
  51. J. M. Robins, “The analysis of randomized and non-randomized aids treatments trials using a new approach to casual inference in longitudinal studies,” in Health Service Research Methodology: A Focus on AIDS (L. Sechrest, H. Freeman, and A. Mulley, eds.), U.S. Public Health Service, Washington D.C., pp. 113-159, 1989a.Google Scholar
  52. J. M. Robins and S. Greenland, “The probability of causation under a stochastic model for individual risk,” Biometrics, 45, pp. 1125-1138, 1989b.Google Scholar
  53. J. M. Robins and S. Greenland, “Identifiability and exchangeability for direct and indirect effects,” Epidemiology, 3(2), pp. 143-155, 1992.Google Scholar
  54. J. M. Robins, “Anewapproach to causal inference in mortality studies with a sustained exposure period-applications to control of the healthy workers survivor effect,” Mathematical Modeling, 7, pp. 1393-1512, 1986.Google Scholar
  55. J. M. Robins, “A graphical approach to the identification and estimation of causal parameters in mortality studies with sustained exposure periods,” Journal of Chronic Diseases, 40(Suppl 2), pp. 139S-161S, 1987.Google Scholar
  56. J. M. Robins, “Data, design, and background knowledge in etiologic inference,” Epidemiology, 12(3), pp. 313-320, 2001.Google Scholar
  57. P. Rosenbaum and D. Rubin, “The central role of propensity score in observational studies for causal effects,” Biometrica, 70, pp. 41-55, 1983.Google Scholar
  58. D. B. Rubin, “Estimating causal effects of treatments in randomized and nonrandomized studies,” Journal of Educational Psychology, 66, pp. 688-701, 1974.Google Scholar
  59. H. A. Simon and N. Rescher, “Cause and counterfactual,” Philosophy and Science, 33, pp. 323-340, 1966.Google Scholar
  60. H. A. Simon, “Causal ordering and identifiability,” in Studies in Econometric Method (Wm. C. Hood and T. C. Koopmans, eds.), Wiley and Sons Inc., pp. 49-74, 1953.Google Scholar
  61. M. E. Sobel, “Causal inference in statistical models of the process of socioeconomic achievement,” Sociological Methods & Research, 27(2), pp. 318-348, 1998.Google Scholar
  62. P. Spirtes, C. Glymour and R. Scheines. Causation, Prediction, and Search, Springer-Verlag, New York, 1993.Google Scholar
  63. P. Suppes. A Probabilistic Theory of Causality, North-Holland Publishing Co., Amsterdam, 1970.Google Scholar
  64. J. Tian and J. Pearl, “Probabilities of causation: Bounds and identification,” in Proceedings of the Sixteenth Conference on Uncertainty in Artificial Intelligence, Morgan Kaufmann, San Francisco, CA, pp. 589-598, 2000.Google Scholar
  65. J. Tian and J. Pearl, “On the identification of causal effects,” in Proceedings of the American Association of Artificial Intelligence, AAAI Press=The MIT Press, Menlo Park, CA, 2002.Google Scholar
  66. J. Tian, A. Paz and J. Pearl, “Finding minimal separating sets,” Technical Report R-254, University of California, Los Angeles, CA, 1998.Google Scholar
  67. C. R. Weinberg, “Toward a clearer definition of confounding,” American Journal of Epidemiology, 137, pp. 1-8, 1993.Google Scholar
  68. N. Wermuth and D. Cox, “Linear dependencies represented by chain graphs,” Statistical Science, 8(3), pp. 204-218, 1993.Google Scholar
  69. N. Wermuth, “On block-recursive regression equations (with discussion),” Brazilian Journal of Probability and Statistics, 6, pp. 1-56, 1992.Google Scholar
  70. J. Whittaker. Graphical Models in Applied Multivariate Statistics, John Wiley, Chichester, England, 1990.Google Scholar
  71. A. S. Whittemore, “Collapsibility of multidimensional contingency tables,” Journal of the Royal Statistical Society, B, 40(3), pp. 328-340, 1978.Google Scholar
  72. S. Wright, “Correlation and causation,” Journal of Agricultural Research, 20, pp. 557-585, 1921.Google Scholar

Copyright information

© Kluwer Academic Publishers 2001

Authors and Affiliations

  • Judea Pearl
    • 1
  1. 1.Cognitive Systems Laboratory, Computer Science DepartmentUniversity of CaliforniaLos Angeles

Personalised recommendations