A Flexible, Interpretable Framework for Assessing Sensitivity to Unmeasured Confounding

Autor/inn/en	Dorie, Vincent; Harada, Masataka; Carnegie, Nicole Bohme; Hill, Jennifer
Titel	A Flexible, Interpretable Framework for Assessing Sensitivity to Unmeasured Confounding
Quelle	35 (2016), S.3453-3470 (18 Seiten)Infoseite zur Zeitschrift PDF als Volltext (1); PDF als Volltext kostenfreie Datei (2) Verfügbarkeit
Zusatzinformation	Weitere Informationen
Sprache	englisch
Dokumenttyp	gedruckt; online; Zeitschriftenaufsatz
DOI	10.1002/sim.6973
Schlagwörter	Bayesian Statistics; Mathematical Models; Causal Models; Statistical Bias; Nonparametric Statistics; Research Problems; Error of Measurement; Statistical Inference; Monte Carlo Methods; Markov Processes; Medicine + Suchen Sie Ihr Suchwort? Mathematical model; Mathematisches Modell; Kausalanalyse; Forschungskritik; Messfehler; Inferential statistics; Schließende Statistik; Monte-Carlo-Methode; Markowscher Prozess; Medizin
Abstract	When estimating causal effects, unmeasured confounding and model misspecification are both potential sources of bias. We propose a method to simultaneously address both issues in the form of a semi-parametric sensitivity analysis. In particular, our approach incorporates Bayesian Additive Regression Trees into a two-parameter sensitivity analysis strategy that assesses sensitivity of posterior distributions of treatment effects to choices of sensitivity parameters. This results in an easily interpretable framework for testing for the impact of an unmeasured confounder that also limits the number of modeling assumptions. We evaluate our approach in a large-scale simulation setting and with high blood pressure data taken from the Third National Health and Nutrition Examination Survey. The model is implemented as open-source software, integrated into the treatSens package for the R statistical programming language. (As Provided).
Erfasst von	ERIC (Education Resources Information Center), Washington, DC
Update	2020/1/01