An alternative robust estimator of average treatment effect in causal inference

Jianxuan Liu, Yanyuan Ma, Lan Wang

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

The problem of estimating the average treatment effects is important when evaluating the effectiveness of medical treatments or social intervention policies. Most of the existing methods for estimating the average treatment effect rely on some parametric assumptions about the propensity score model or the outcome regression model one way or the other. In reality, both models are prone to misspecification, which can have undue influence on the estimated average treatment effect. We propose an alternative robust approach to estimating the average treatment effect based on observational data in the challenging situation when neither a plausible parametric outcome model nor a reliable parametric propensity score model is available. Our estimator can be considered as a robust extension of the popular class of propensity score weighted estimators. This approach has the advantage of being robust, flexible, data adaptive, and it can handle many covariates simultaneously. Adopting a dimension reduction approach, we estimate the propensity score weights semiparametrically by using a non-parametric link function to relate the treatment assignment indicator to a low-dimensional structure of the covariates which are formed typically by several linear combinations of the covariates. We develop a class of consistent estimators for the average treatment effect and study their theoretical properties. We demonstrate the robust performance of the estimators on simulated data and a real data example of investigating the effect of maternal smoking on babies’ birth weight.

Original languageEnglish (US)
Pages (from-to)910-923
Number of pages14
JournalBiometrics
Volume74
Issue number3
DOIs
StatePublished - Sep 2018

Fingerprint

Average Treatment Effect
Causal Inference
Propensity Score
Robust Estimators
Alternatives
Covariates
Estimator
Public Policy
Link Function
Birth Weight
Robust Performance
medical treatment
Misspecification
Consistent Estimator
Smoking
infants
Dimension Reduction
maternal effect
Theoretical Models
birth weight

All Science Journal Classification (ASJC) codes

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

Cite this

Liu, Jianxuan ; Ma, Yanyuan ; Wang, Lan. / An alternative robust estimator of average treatment effect in causal inference. In: Biometrics. 2018 ; Vol. 74, No. 3. pp. 910-923.
@article{942b24bc16764206b1b8ff097ac6e53c,
title = "An alternative robust estimator of average treatment effect in causal inference",
abstract = "The problem of estimating the average treatment effects is important when evaluating the effectiveness of medical treatments or social intervention policies. Most of the existing methods for estimating the average treatment effect rely on some parametric assumptions about the propensity score model or the outcome regression model one way or the other. In reality, both models are prone to misspecification, which can have undue influence on the estimated average treatment effect. We propose an alternative robust approach to estimating the average treatment effect based on observational data in the challenging situation when neither a plausible parametric outcome model nor a reliable parametric propensity score model is available. Our estimator can be considered as a robust extension of the popular class of propensity score weighted estimators. This approach has the advantage of being robust, flexible, data adaptive, and it can handle many covariates simultaneously. Adopting a dimension reduction approach, we estimate the propensity score weights semiparametrically by using a non-parametric link function to relate the treatment assignment indicator to a low-dimensional structure of the covariates which are formed typically by several linear combinations of the covariates. We develop a class of consistent estimators for the average treatment effect and study their theoretical properties. We demonstrate the robust performance of the estimators on simulated data and a real data example of investigating the effect of maternal smoking on babies’ birth weight.",
author = "Jianxuan Liu and Yanyuan Ma and Lan Wang",
year = "2018",
month = "9",
doi = "10.1111/biom.12859",
language = "English (US)",
volume = "74",
pages = "910--923",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "3",

}

An alternative robust estimator of average treatment effect in causal inference. / Liu, Jianxuan; Ma, Yanyuan; Wang, Lan.

In: Biometrics, Vol. 74, No. 3, 09.2018, p. 910-923.

Research output: Contribution to journalArticle

TY - JOUR

T1 - An alternative robust estimator of average treatment effect in causal inference

AU - Liu, Jianxuan

AU - Ma, Yanyuan

AU - Wang, Lan

PY - 2018/9

Y1 - 2018/9

N2 - The problem of estimating the average treatment effects is important when evaluating the effectiveness of medical treatments or social intervention policies. Most of the existing methods for estimating the average treatment effect rely on some parametric assumptions about the propensity score model or the outcome regression model one way or the other. In reality, both models are prone to misspecification, which can have undue influence on the estimated average treatment effect. We propose an alternative robust approach to estimating the average treatment effect based on observational data in the challenging situation when neither a plausible parametric outcome model nor a reliable parametric propensity score model is available. Our estimator can be considered as a robust extension of the popular class of propensity score weighted estimators. This approach has the advantage of being robust, flexible, data adaptive, and it can handle many covariates simultaneously. Adopting a dimension reduction approach, we estimate the propensity score weights semiparametrically by using a non-parametric link function to relate the treatment assignment indicator to a low-dimensional structure of the covariates which are formed typically by several linear combinations of the covariates. We develop a class of consistent estimators for the average treatment effect and study their theoretical properties. We demonstrate the robust performance of the estimators on simulated data and a real data example of investigating the effect of maternal smoking on babies’ birth weight.

AB - The problem of estimating the average treatment effects is important when evaluating the effectiveness of medical treatments or social intervention policies. Most of the existing methods for estimating the average treatment effect rely on some parametric assumptions about the propensity score model or the outcome regression model one way or the other. In reality, both models are prone to misspecification, which can have undue influence on the estimated average treatment effect. We propose an alternative robust approach to estimating the average treatment effect based on observational data in the challenging situation when neither a plausible parametric outcome model nor a reliable parametric propensity score model is available. Our estimator can be considered as a robust extension of the popular class of propensity score weighted estimators. This approach has the advantage of being robust, flexible, data adaptive, and it can handle many covariates simultaneously. Adopting a dimension reduction approach, we estimate the propensity score weights semiparametrically by using a non-parametric link function to relate the treatment assignment indicator to a low-dimensional structure of the covariates which are formed typically by several linear combinations of the covariates. We develop a class of consistent estimators for the average treatment effect and study their theoretical properties. We demonstrate the robust performance of the estimators on simulated data and a real data example of investigating the effect of maternal smoking on babies’ birth weight.

UR - http://www.scopus.com/inward/record.url?scp=85041809241&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85041809241&partnerID=8YFLogxK

U2 - 10.1111/biom.12859

DO - 10.1111/biom.12859

M3 - Article

C2 - 29441521

AN - SCOPUS:85041809241

VL - 74

SP - 910

EP - 923

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 3

ER -