Bayesian Inference for the Causal Effect of Mediation

Michael J. Daniels, Jason Roy, Chanmin Kim, Joseph W. Hogan, Michael G. Perri

Research output: Contribution to journalArticle

20 Citations (Scopus)

Abstract

We propose a nonparametric Bayesian approach to estimate the natural direct and indirect effects through a mediator in the setting of a continuous mediator and a binary response. Several conditional independence assumptions are introduced (with corresponding sensitivity parameters) to make these effects identifiable from the observed data. We suggest strategies for eliciting sensitivity parameters and conduct simulations to assess violations to the assumptions. This approach is used to assess mediation in a recent weight management clinical trial.

Original languageEnglish (US)
Pages (from-to)1028-1036
Number of pages9
JournalBiometrics
Volume68
Issue number4
DOIs
StatePublished - Dec 1 2012
Externally publishedYes

Fingerprint

Parameter Sensitivity
Causal Effect
Bayes Theorem
Mediation
Mediator
weight control
Bayesian inference
clinical trials
Clinical Trials
Weights and Measures
Binary Response
Conditional Independence
Bayesian Approach
Estimate
Simulation
Strategy

All Science Journal Classification (ASJC) codes

  • Applied Mathematics
  • Statistics and Probability
  • Agricultural and Biological Sciences(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Immunology and Microbiology(all)
  • Medicine(all)

Cite this

Daniels, M. J., Roy, J., Kim, C., Hogan, J. W., & Perri, M. G. (2012). Bayesian Inference for the Causal Effect of Mediation. Biometrics, 68(4), 1028-1036. https://doi.org/10.1111/j.1541-0420.2012.01781.x
Daniels, Michael J. ; Roy, Jason ; Kim, Chanmin ; Hogan, Joseph W. ; Perri, Michael G. / Bayesian Inference for the Causal Effect of Mediation. In: Biometrics. 2012 ; Vol. 68, No. 4. pp. 1028-1036.
@article{475f8638dfd6423da3c317749723df3d,
title = "Bayesian Inference for the Causal Effect of Mediation",
abstract = "We propose a nonparametric Bayesian approach to estimate the natural direct and indirect effects through a mediator in the setting of a continuous mediator and a binary response. Several conditional independence assumptions are introduced (with corresponding sensitivity parameters) to make these effects identifiable from the observed data. We suggest strategies for eliciting sensitivity parameters and conduct simulations to assess violations to the assumptions. This approach is used to assess mediation in a recent weight management clinical trial.",
author = "Daniels, {Michael J.} and Jason Roy and Chanmin Kim and Hogan, {Joseph W.} and Perri, {Michael G.}",
year = "2012",
month = "12",
day = "1",
doi = "10.1111/j.1541-0420.2012.01781.x",
language = "English (US)",
volume = "68",
pages = "1028--1036",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "4",

}

Daniels, MJ, Roy, J, Kim, C, Hogan, JW & Perri, MG 2012, 'Bayesian Inference for the Causal Effect of Mediation', Biometrics, vol. 68, no. 4, pp. 1028-1036. https://doi.org/10.1111/j.1541-0420.2012.01781.x

Bayesian Inference for the Causal Effect of Mediation. / Daniels, Michael J.; Roy, Jason; Kim, Chanmin; Hogan, Joseph W.; Perri, Michael G.

In: Biometrics, Vol. 68, No. 4, 01.12.2012, p. 1028-1036.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Bayesian Inference for the Causal Effect of Mediation

AU - Daniels, Michael J.

AU - Roy, Jason

AU - Kim, Chanmin

AU - Hogan, Joseph W.

AU - Perri, Michael G.

PY - 2012/12/1

Y1 - 2012/12/1

N2 - We propose a nonparametric Bayesian approach to estimate the natural direct and indirect effects through a mediator in the setting of a continuous mediator and a binary response. Several conditional independence assumptions are introduced (with corresponding sensitivity parameters) to make these effects identifiable from the observed data. We suggest strategies for eliciting sensitivity parameters and conduct simulations to assess violations to the assumptions. This approach is used to assess mediation in a recent weight management clinical trial.

AB - We propose a nonparametric Bayesian approach to estimate the natural direct and indirect effects through a mediator in the setting of a continuous mediator and a binary response. Several conditional independence assumptions are introduced (with corresponding sensitivity parameters) to make these effects identifiable from the observed data. We suggest strategies for eliciting sensitivity parameters and conduct simulations to assess violations to the assumptions. This approach is used to assess mediation in a recent weight management clinical trial.

UR - http://www.scopus.com/inward/record.url?scp=84871667740&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84871667740&partnerID=8YFLogxK

U2 - 10.1111/j.1541-0420.2012.01781.x

DO - 10.1111/j.1541-0420.2012.01781.x

M3 - Article

C2 - 23005030

AN - SCOPUS:84871667740

VL - 68

SP - 1028

EP - 1036

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 4

ER -