Sensitivity analysis for causal inference using inverse probability weighting

Changyu Shen, Xiaochun Li, Lingling Li, Martin C. Were

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

Evaluation of impact of potential uncontrolled confounding is an important component for causal inference based on observational studies. In this article, we introduce a general framework of sensitivity analysis that is based on inverse probability weighting. We propose a general methodology that allows both non-parametric and parametric analyses, which are driven by two parameters that govern the magnitude of the variation of the multiplicative errors of the propensity score and their correlations with the potential outcomes. We also introduce a specific parametric model that offers a mechanistic view on how the uncontrolled confounding may bias the inference through these parameters. Our method can be readily applied to both binary and continuous outcomes and depends on the covariates only through the propensity score that can be estimated by any parametric or non-parametric method. We illustrate our method with two medical data sets.

Original languageEnglish
Pages (from-to)822-837
Number of pages16
JournalBiometrical Journal
Volume53
Issue number5
DOIs
StatePublished - Sep 2011

Fingerprint

Inverse Probability Weighting
Propensity Score
Causal Inference
Confounding
Sensitivity Analysis
Potential Outcomes
Observational Study
Nonparametric Methods
Parametric Model
Covariates
Two Parameters
Multiplicative
Binary
Observational Studies
Methodology
Evaluation
Inverse probability weighting
Propensity score
Causal inference
Sensitivity analysis

Keywords

  • Causal inference
  • Inverse probability weighting
  • Propensity score
  • Sensitivity analysis
  • Uncontrolled confounding

ASJC Scopus subject areas

  • Statistics and Probability
  • Medicine(all)
  • Statistics, Probability and Uncertainty

Cite this

Sensitivity analysis for causal inference using inverse probability weighting. / Shen, Changyu; Li, Xiaochun; Li, Lingling; Were, Martin C.

In: Biometrical Journal, Vol. 53, No. 5, 09.2011, p. 822-837.

Research output: Contribution to journalArticle

Shen, Changyu ; Li, Xiaochun ; Li, Lingling ; Were, Martin C. / Sensitivity analysis for causal inference using inverse probability weighting. In: Biometrical Journal. 2011 ; Vol. 53, No. 5. pp. 822-837.
@article{97484c7a45314468a5dcb755d38c5d6b,
title = "Sensitivity analysis for causal inference using inverse probability weighting",
abstract = "Evaluation of impact of potential uncontrolled confounding is an important component for causal inference based on observational studies. In this article, we introduce a general framework of sensitivity analysis that is based on inverse probability weighting. We propose a general methodology that allows both non-parametric and parametric analyses, which are driven by two parameters that govern the magnitude of the variation of the multiplicative errors of the propensity score and their correlations with the potential outcomes. We also introduce a specific parametric model that offers a mechanistic view on how the uncontrolled confounding may bias the inference through these parameters. Our method can be readily applied to both binary and continuous outcomes and depends on the covariates only through the propensity score that can be estimated by any parametric or non-parametric method. We illustrate our method with two medical data sets.",
keywords = "Causal inference, Inverse probability weighting, Propensity score, Sensitivity analysis, Uncontrolled confounding",
author = "Changyu Shen and Xiaochun Li and Lingling Li and Were, {Martin C.}",
year = "2011",
month = "9",
doi = "10.1002/bimj.201100042",
language = "English",
volume = "53",
pages = "822--837",
journal = "Biometrical Journal",
issn = "0323-3847",
publisher = "Wiley-VCH Verlag",
number = "5",

}

TY - JOUR

T1 - Sensitivity analysis for causal inference using inverse probability weighting

AU - Shen, Changyu

AU - Li, Xiaochun

AU - Li, Lingling

AU - Were, Martin C.

PY - 2011/9

Y1 - 2011/9

N2 - Evaluation of impact of potential uncontrolled confounding is an important component for causal inference based on observational studies. In this article, we introduce a general framework of sensitivity analysis that is based on inverse probability weighting. We propose a general methodology that allows both non-parametric and parametric analyses, which are driven by two parameters that govern the magnitude of the variation of the multiplicative errors of the propensity score and their correlations with the potential outcomes. We also introduce a specific parametric model that offers a mechanistic view on how the uncontrolled confounding may bias the inference through these parameters. Our method can be readily applied to both binary and continuous outcomes and depends on the covariates only through the propensity score that can be estimated by any parametric or non-parametric method. We illustrate our method with two medical data sets.

AB - Evaluation of impact of potential uncontrolled confounding is an important component for causal inference based on observational studies. In this article, we introduce a general framework of sensitivity analysis that is based on inverse probability weighting. We propose a general methodology that allows both non-parametric and parametric analyses, which are driven by two parameters that govern the magnitude of the variation of the multiplicative errors of the propensity score and their correlations with the potential outcomes. We also introduce a specific parametric model that offers a mechanistic view on how the uncontrolled confounding may bias the inference through these parameters. Our method can be readily applied to both binary and continuous outcomes and depends on the covariates only through the propensity score that can be estimated by any parametric or non-parametric method. We illustrate our method with two medical data sets.

KW - Causal inference

KW - Inverse probability weighting

KW - Propensity score

KW - Sensitivity analysis

KW - Uncontrolled confounding

UR - http://www.scopus.com/inward/record.url?scp=80052404588&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80052404588&partnerID=8YFLogxK

U2 - 10.1002/bimj.201100042

DO - 10.1002/bimj.201100042

M3 - Article

VL - 53

SP - 822

EP - 837

JO - Biometrical Journal

JF - Biometrical Journal

SN - 0323-3847

IS - 5

ER -