New Perspectives on Causal Mediation Analysis

  • Xiaolu Wang
  • Michael E. Sobel
Part of the Handbooks of Sociology and Social Research book series (HSSR)


Social and behavioral scientists have long used path analysis and related linear structural equation models (SEMs) to decompose parameters of ordered systems of equations into “direct effects” and “indirect effects” through mediating variables. These decompositions have been used to address substantive questions of fundamental interest, for example, how a person’s social background affects his/her earnings through education. However, in general, the “direct effects” and “indirect effects” defined in and estimated from these models should not be given causal interpretations, even in randomized experiments. To illustrate this, we first define various direct and indirect effects using potential outcomes notation and discuss situations where an investigator might want to consider these. Second, we consider identification of these effects: the required identifying assumptions are more often than not implausible for the kinds of data collected and questions considered in social and behavioral research. Third, we present other identifying assumptions that might be used to identify direct and indirect effects, and then briefly discuss different methods to estimate these effects, including regression, instrumental variables, marginal structural models, and weighting methods.Finally, we introduce an alternative approach to mediation (principal stratification), define several possible effects of interest, and briefly discuss identifying assumptions and estimation.


Structural Equation Model Causal Model Control Effect Marginal Structural Model Causal Parameter 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.



We thank Felix Elwert, Stephen Morgan, Geoffrey Wodtke, Kenneth Bollen and Judea Pearl for helpful comments. Any remaining errors are the authors’ alone.


  1. Angrist, J. D., Imbens, G. W., & Rubin, D. B. (1996). Identification of causal effects using instrumental variables. Journal of the American Statistical Association, 91, 444–472.CrossRefGoogle Scholar
  2. Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182.CrossRefGoogle Scholar
  3. Belsen, W. A. (1956). A technique for studying the effects of a television broadcast. Applied Statistics, 5, 195–202.CrossRefGoogle Scholar
  4. Bolger, N., & Amarel, D. (2007). Effects of social support visibility on adjustment to stress: Experimental evidence. Journal of Personality and Social Psychology, 92, 458–475.CrossRefGoogle Scholar
  5. Brumback, B. A., Hernán, M. A., Haneuse, S. J., & Robins, J. M. (2004). Sensitivity analyses for unmeasured confounding assuming a marginal structural model for repeated measures. Statistics in Medicine, 23, 749–767.CrossRefGoogle Scholar
  6. Dawid, A. P. (2008). Beware of the DAG!. Journal of Machine Learning Research: Workshop and Conference Proceeding, 6, 59–86.Google Scholar
  7. Duncan, O. D. (1966). Path analysis: Sociological examples. American Journal of Sociology, 76, 1–16.CrossRefGoogle Scholar
  8. Emsley, R., Dunn, G., & White, I. R. (2010). Mediation and moderation of treatment effects in randomised controlled trials of complex interventions. Statistical Methods in Medical Research, 19, 237–270.CrossRefGoogle Scholar
  9. Frangakis, C. E., & Rubin, D. B. (1999). Addressing complications of intentto-treat analysis in the combined presence of all-or-none treatment-noncompliance and subsequent missing outcomes. Biometrika, 86, 365–379.CrossRefGoogle Scholar
  10. Frangakis, C. E., & Rubin, D. B. (2002). Principal stratification in causal inference. Biometrics, 58, 21–29.CrossRefGoogle Scholar
  11. Gennetian, L. A., Morris, P. A., Bos, J. M., & Bloom, H. S. (2005). Constructing instrumental variables from experimental data to explore how treatments produce effects. In H. S. Bloom (Ed.), Learning more from social experiments: Evolving analytic approaches (1st ed., pp. 75–114). New York: Russell Sage Foundation.Google Scholar
  12. Halaby, C. N. (1979). Job-specific sex differences in organizational reward attainment: Wage discrimination vs. rank segregation. Social Forces, 58, 108–127.Google Scholar
  13. Hauser, R. M., & Featherman, D. L. (1977). The process of stratification: Trends and analyses. New York: Academic Press.Google Scholar
  14. Hernán, M. A., & Robins, J. M. (2009). Estimation of the causal effects of time-varying exposure. In G. Fitzmaurice, M. Davidian, G. Verbeke, & G. Molenberghs (Eds.), Longitudinal data analysis. Boca Raton: Chapman and Hall/CRC.Google Scholar
  15. Hill, J., Waldfogel, J., & Brooks-Gunn, J. (2002). Differential effects of high-quality child care. Journal of Policy Analysis and Management, 21, 601–627.CrossRefGoogle Scholar
  16. Hirano, K., Imbens, G. W., Rubin, D. B., & Zhou, X.-H. (2000). Assessing the effect of an influenza vaccine in an encouragement design. Biostatistics, 1, 69–88.CrossRefGoogle Scholar
  17. Holland, P. W. (1988). Causal inference, path analysis, and recursive structural equation models (with discussion). Sociological Methodology, 18, 449–493.CrossRefGoogle Scholar
  18. Hong, G., & Nomi, T. (2012). Weighting methods for assessing policy effects mediated by peer change. Journal of Educational Effectiveness (special issue on the statistical approaches to studying mediator effects in education research), 5, 261–289.Google Scholar
  19. Horvitz, D. G., & Thompson, D. J. (1952). A generalization of sampling without replacement from a finite universe. Journal of the American Statistical Association, 47, 663–685.CrossRefGoogle Scholar
  20. Imai, K., Keele, L., & Yamamoto, T. (2010). Identification, inference and sensitivity analysis for causal mediation effects. Statistical Science, 25, 51–71.CrossRefGoogle Scholar
  21. Imai, K., Tingley, D., & Yamamoto, T. (2013). Experimental designs for identifying causal mechanisms. Journal of the Royal Statistical Society, Series A, 176, 5–51.Google Scholar
  22. Jin, H., & Rubin, D. B. (2008). Principal stratification for causal inference with extended partial compliance. Journal of the American Statistical Association, 103, 101–111.CrossRefGoogle Scholar
  23. Jin, H., & Rubin, D. B. (2009). Public schools versus private schools: Causal inference with partial compliance. Journal of Educational and Behavioral Statistics, 34, 24–45.CrossRefGoogle Scholar
  24. Jo, B. (2002). Estimating intervention effects with noncompliance: Alternative model specifications. Journal of Educational and Behavioral Statistics, 27, 385–430.CrossRefGoogle Scholar
  25. Jo, B. (2008). Causal inference in randomized experiments with mediational processes. Psychological Methods, 13, 314–336.CrossRefGoogle Scholar
  26. Jo, B., & Stuart, E. A. (2009). On the use of propensity scores in principal causal effect estimation. Statistics in Medicine, 28, 2857–2875.CrossRefGoogle Scholar
  27. Joffe, M. M., Small, D., & Hsu, C.-Y. (2007). Defining and estimating intervention effects for groups that will develop an auxiliary outcome. Statistical Science, 22, 74–97.CrossRefGoogle Scholar
  28. Jöreskog, K. (1977). Structural equation models in the social sciences: Specification, estimation and testing. In P. R. Krishnaiah (Ed.), Application of Statistics (pp. 265–287). Amsterdam: North-Holland.Google Scholar
  29. Lindquist, M. A. (2012). Functional causal mediation analysis with an application to brain connectivity. Journal of the American Statistical Association, 107, 1297–1309.CrossRefGoogle Scholar
  30. Lindquist, M. A., & Sobel, M. E. (2011a). Graphical models, potential outcomes and causal inference: Comment on Ramsey, Spirtes and Glymour. Neuroimage, 57, 334–336.CrossRefGoogle Scholar
  31. Lindquist, M. A., & Sobel, M. E. (2011b). Cloak and DAG: A response to the comments on our comment. Neuroimage, doi:10.1016/j.neuroimage.2011.11.027.Google Scholar
  32. Little, R. J., & Yau, L. H. Y. (1998). Statistical techniques for analyzing data from prevention trials: Treatment of no-shows using Rubin’s causal model. Psychological Methods, 2, 147–159.CrossRefGoogle Scholar
  33. MacKinnon, D. P. (2008). Introduction to statistical mediation analysis. New York: Lawrence Erlbaum Associates.Google Scholar
  34. Molm, L. D., Takahashi, N., & Peterson, G. (2003). In the eye of the beholder: Procedural justice in social exchange. American Sociological Review, 68, 128–152.CrossRefGoogle Scholar
  35. Pearl, J. (2001). Direct and indirect effects. In Proceedings of the seventeenth conference on uncertainty and artificial intelligence (pp. 411–420). San Francisco, CA: Morgan Kaufmann.Google Scholar
  36. Pearl, J. (2009). Causality: Models, reasoning, and inference. Cambridge: Cambridge University Press.CrossRefGoogle Scholar
  37. Peterson, M. L., Sinisi, S. E., & van der Laan, M. J. (2006). Estimation of direct causal effects. Epidemiology, 17, 276–284.CrossRefGoogle Scholar
  38. Riach, P. A., & Rich, J. (2006). An experimental investigation of sexual discrimination and hiring in the English labor market. The B. E. Journal of Economic Analysis and Policy, 6(2), 1.Google Scholar
  39. Robins, J. M. (1986). A new approach to causal inference in mortality studies with sustained exposure periods – Application to control of the healthy worker survivor effect. Mathematical Modeling, 7, 1393–1512.CrossRefGoogle Scholar
  40. Robins, J. M. (1999). Association, causation and marginal structural models. Synthese, 121, 51–179.CrossRefGoogle Scholar
  41. Robins, J. M. (2003). Semantics of causal DAG models and the identification of direct and indirect effects. In P. Green, N. L. Hjort, & S. Richardson (Eds.), Highly structured stochastic systems (pp. 70–81). New York: Oxford University Press.Google Scholar
  42. Robins, J. M., & Greenland, S. (1992). Identifiability and exchangeability of direct and indirect effects. Epidemiology, 3, 143–155.CrossRefGoogle Scholar
  43. Rosenbaum, P. R. (1984). The consequences of adjustment for a concomitant variable that has been affected by the treatment. Journal of the Royal Statistical Society, Series A, 147, 656–666 (General).Google Scholar
  44. Rosenbaum, P. R. (2010). Design of observational studies. New York: Springer.CrossRefGoogle Scholar
  45. Rubin, D. B. (1977). Assignment to treatment group on the basis of a covariate. Journal of Educational Statistics, 2, 1–26.CrossRefGoogle Scholar
  46. Rubin, D. B. (1980). Discussion of ‘randomization analysis of experimental data: The Fisher randomization test’ by D. Basu. Journal of the American Statistical Association, 75, 591–593.Google Scholar
  47. Small, D., Ten Have, T., Joe, M., & Cheng, J. (2006). Random effects logistic models for analysing efficacy of a longitudinal randomized treatment with non-adherence. Statistics in Medicine, 25, 1981–2007.CrossRefGoogle Scholar
  48. Sobel, M. E. (1990). Effect analysis and causation in linear structural equation models. Psychometrika, 55, 495–515.CrossRefGoogle Scholar
  49. Sobel, M. E. (2008). Identification of causal parameters in randomized studies with mediating variables. Journal of Educational and Behavioral Statistics, 33, 230–251.CrossRefGoogle Scholar
  50. Sobel, M. E. (2012). Does marriage boost men’s wages?: Identification of treatment effects in fixed and random effects regression models for panel data. Journal of the American Statistical Association, 107, 521–529.CrossRefGoogle Scholar
  51. Sobel, M. E., & Muthén, B. O. (2012). Compliance mixture modeling with a zero effect complier class and missing data. Biometrics, 68, 1037–1045.CrossRefGoogle Scholar
  52. VanderWeele, T. J. (2009). Marginal structural models for the estimation of direct and indirect effects. Epidemiology, 20, 18–26.CrossRefGoogle Scholar
  53. VanderWeele, T. J. (2010). Bias formulas for sensitivity analysis for direct and indirect effects. Epidemiology, 21, 540–551.CrossRefGoogle Scholar
  54. VanderWeele, T. J., & Vansteelandt, S. (2009). Conceptual issues concerning mediation, interventions and composition. Statistics and Its Interface, 2, 457–468.Google Scholar
  55. Yau, L. H. Y., & Little, R. J. (2001). Inference for the complier-average causal effect from longitudinal data subject to noncompliance and missing data, with application to a job training assessment for the unemployed. Journal of the American Statistical Association, 96, 1232–1244.CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2013

Authors and Affiliations

  • Xiaolu Wang
    • 1
  • Michael E. Sobel
    • 2
  1. 1.Department of SociologyColumbia UniversityNew YorkUSA
  2. 2.Department of StatisticsColumbia UniversityNew YorkUSA

Personalised recommendations