Sharp sensitivity bounds for mediation under unmeasured mediator-outcome confounding

被引：51

作者：

Ding, Peng ^{[1
]}

Vanderweele, Tyler J. ^{[2
]}

机构：

[1] Univ Calif Berkeley, Dept Stat, Berkeley, CA 94720 USA

[2] Harvard Univ, TH Chan Sch Publ Hlth, Dept Epidemiol, Boston, MA 02115 USA

来源：

BIOMETRIKA | 2016年 / 103卷 / 02期

基金：

美国国家卫生研究院;

关键词：

Bounding factor; Causal inference; Collider; Natural direct effect; Natural indirect effect; NATURAL DIRECT; LUNG-CANCER; IDENTIFICATION; SMOKING; MODELS;

D O I：

10.1093/biomet/asw012

中图分类号：

Q [生物科学];

学科分类号：

07 ; 0710 ; 09 ;

摘要：

It is often of interest to decompose the total effect of an exposure into a component that acts on the outcome through some mediator and a component that acts independently through other pathways. Said another way, we are interested in the direct and indirect effects of the exposure on the outcome. Even if the exposure is randomly assigned, it is often infeasible to randomize the mediator, leaving the mediator-outcome confounding not fully controlled. We develop a sensitivity analysis technique that can bound the direct and indirect effects without parametric assumptions about the unmeasured mediator-outcome confounding.

引用

页码：483 / 490

页数：8

共 24 条

[1]

[Anonymous], 2015, EXPLANATION CAUSAL I

[2]

[Anonymous], 2009, CAUSALITY MODELS REA

[3] THE MODERATOR MEDIATOR VARIABLE DISTINCTION IN SOCIAL PSYCHOLOGICAL-RESEARCH - CONCEPTUAL, STRATEGIC, AND STATISTICAL CONSIDERATIONS [J].

BARON, RM ;

KENNY, DA .

JOURNAL OF PERSONALITY AND SOCIAL PSYCHOLOGY, 1986, 51 (06) :1173-1182

[4]

CORNFIELD J, 1959, JNCI-J NATL CANCER I, V22, P173

[5]

Didelez V., 2006, Proceedings of the Twenty-Second Conference on Uncertainty in Artificial Intelligence, P138

[6] Generalized Cornfield conditions for the risk difference [J].

Ding, Peng ;

Vanderweele, Tyler J. .

BIOMETRIKA, 2014, 101 (04) :971-977

[7] Identifying direct and indirect effects in a non-counterfactual framework [J].

Geneletti, Sara .

JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY, 2007, 69 :199-215

[8] Quantifying biases in causal models:: Classical confounding vs collider-stratification bias [J].

Greenland, S .

EPIDEMIOLOGY, 2003, 14 (03) :300-306

[9] Identification, Inference and Sensitivity Analysis for Causal Mediation Effects [J].

Imai, Kosuke ;

Keele, Luke ;

Yamamoto, Teppei .

STATISTICAL SCIENCE, 2010, 25 (01) :51-71

[10] Identification and Efficient Estimation of the Natural Direct Effect among the Untreated [J].

Lendle, Samuel D. ;

Subbaraman, Meenakshi S. ;

van der Laan, Mark J. .

BIOMETRICS, 2013, 69 (02) :310-317

← 1 2 3 →