Applications of Causally Defined Direct and Indirect Effects in Mediation Analysis using SEM in Mplus

This paper summarizes some of the literature on causal effects in mediation analysis. It presents causally-defined direct and indirect effects for continuous, binary, ordinal, nominal, and count variables. The expansion to non-continuous mediators and outcomes offers a broader array of causal mediation analyses than previously considered in structural equation modeling practice. A new result is the ability to handle mediation by a nominal variable. Examples with a binary outcome and a binary, ordinal or nominal mediator are given using Mplus to compute the effects. The causal effects require strong assumptions even in randomized designs, especially sequential ignorability, which is presumably often violated to some extent due to mediator-outcome confounding. To study the effects of violating this assumption, it is shown how a sensitivity analysis can be carried out. This can be used both in planning a new study and in evaluating the results of an existing study.

[1]  Judea Pearl.  Graphical models, potential outcomes and causal inference: Comment on Linquist and Sobel , 2011, NeuroImage.

[2]  A. Goldberger,et al.  Structural Equation Models in the Social Sciences. , 1974 .

[3]  B. Muthén,et al.  How to Use a Monte Carlo Study to Decide on Sample Size and Determine Power , 2002 .

[4]  D. A. Kenny,et al.  Process Analysis , 1981 .

[5]  L. James,et al.  Mediators, Moderators, and Tests for Mediation. , 1984 .

[6]  Clark Glymour,et al.  Counterfactuals, graphical causal models and potential outcomes: Response to Lindquist and Sobel , 2013, NeuroImage.

[7]  B. Muthén,et al.  Growth mixture modeling , 2008 .

[8]  D P MacKinnon,et al.  The intermediate endpoint effect in logistic and probit regression , 2007, Clinical trials.

[9]  Judea Pearl,et al.  Direct and Indirect Effects , 2001, UAI.

[10]  J. Robins,et al.  Identifiability and Exchangeability for Direct and Indirect Effects , 1992, Epidemiology.

[11]  D. A. Kenny,et al.  The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. , 1986, Journal of personality and social psychology.

[12]  P. Holland.  Causal Inference, Path Analysis and Recursive Structural Equations Models. Program Statistics Research, Technical Report No. 88-81. , 1988 .

[13]  Martin A. Lindquist,et al.  Graphical models, potential outcomes and causal inference: Comment on Ramsey, Spirtes and Glymour , 2011, NeuroImage.

[14]  J. Pearl.  3. The Foundations of Causal Inference , 2010 .

[15]  Helena Chmura Kraemer,et al.  How and why criteria defining moderators and mediators differ between the Baron & Kenny and MacArthur approaches. , 2008, Health psychology : official journal of the Division of Health Psychology, American Psychological Association.

[16]  Christopher Winship,et al.  Structural Equations and Path Analysis for Discrete Data , 1983, American Journal of Sociology.

[17]  Martin A. Lindquist,et al.  Cloak and DAG: A response to the comments on our comment , 2013, NeuroImage.

[18]  Tyler J. VanderWeele,et al.  Conceptual issues concerning mediation, interventions and composition , 2009 .

[19]  John G Bullock,et al.  Yes, But What's the Mechanism? (Don't Expect an Easy Answer) , 2010, Journal of personality and social psychology.

[20]  Jay Magidson,et al.  Advances in factor analysis and structural equation models , 1980 .

[21]  S. Raudenbush,et al.  Evaluating Kindergarten Retention Policy , 2006 .

[22]  Kosuke Imai,et al.  Experimental Designs for Identifying Causal Mechanisms (with discussions) , 2013 .

[23]  A. Philip Dawid,et al.  Causality : statistical perspectives and applications , 2012 .

[24]  B. Muthén.  A general structural equation model with dichotomous, ordered categorical, and continuous latent variable indicators , 1984 .

[25]  J. Pearl.  The Mediation Formula: A Guide to the Assessment of Causal Pathways in Nonlinear Models , 2011 .

[26]  Stijn Vansteelandt,et al.  Odds ratios for mediation analysis for a dichotomous outcome. , 2010, American journal of epidemiology.

[27]  T J VanderWeele,et al.  Direct and Indirect Effects for Neighborhood-Based Clustered and Longitudinal Data , 2010, Sociological methods & research.

[28]  G. Richardson,et al.  Alcohol, marijuana, and tobacco: effects of prenatal exposure on offspring growth and morphology at age six. , 1994, Alcoholism, clinical and experimental research.

[29]  L. Keele,et al.  A General Approach to Causal Mediation Analysis , 2010, Psychological methods.

[30]  A. Satorra.  Structural Equation Models with Latent Variables , 2002 .

[31]  W. Stallings,et al.  SIXTH EDITION , 2000 .

[32]  Bengt Muthén,et al.  A Structural Probit Model with Latent Variables , 1979 .

[33]  M. Sobel.  Identification of Causal Parameters in Randomized Studies With Mediating Variables , 2008 .

[34]  J. Pearl.  The Causal Mediation Formula—A Guide to the Assessment of Pathways and Mechanisms , 2012, Prevention Science.

[35]  D. Mackinnon.  Introduction to Statistical Mediation Analysis , 2008 .

[36]  Kristopher J Preacher,et al.  Addressing Moderated Mediation Hypotheses: Theory, Methods, and Prescriptions , 2007, Multivariate behavioral research.

[37]  Katherine E. Masyn,et al.  General growth mixture modeling for randomized preventive interventions. , 2001, Biostatistics.

[38]  L. Keele,et al.  Identification, Inference and Sensitivity Analysis for Causal Mediation Effects , 2010, 1011.1079.

[39]  J. Ghosh.  Causality: Models, Reasoning and Inference, Second Edition by Judea Pearl , 2011 .

[40]  T. VanderWeele.  Bias Formulas for Sensitivity Analysis for Direct and Indirect Effects , 2010, Epidemiology.