ate: Doubly robust estimation of average treatment effect

In ehkennedy/npcausal: Nonparametric causal inference methods

Description

ate is used to estimate the mean outcome in a population had all subjects received given levels of a discrete (unconfounded) treatment, using doubly robust methods with ensembled nuisance estimation.

Usage

ate(y, a, x, nsplits=2, sl.lib=c("SL.earth","SL.gam","SL.glm","SL.glmnet",
  "SL.glm.interaction", "SL.mean","SL.ranger","rpart"))

Arguments

y	outcome of interest.
a	discrete treatment.
x	covariate matrix.
nsplits	integer number of sample splits for nuisance estimation. If nsplits=1, sample splitting is not used, and nuisance functions are estimated on full sample (in which case validity of SEs/CIs requires empirical process conditions). Otherwise must have nsplits>1.
sl.lib	algorithm library for SuperLearner. Default library includes "earth", "gam", "glm", "glmnet", "glm.interaction", "mean", "ranger", "rpart.

Value

A list containing the following components:

res	estimates/SEs/CIs/p-values for population means and relevant contrasts.
nuis	subject-specific estimates of nuisance functions (i.e., propensity score and outcome regression)
ifvals	matrix of estimated influence function values.

References

Robins JM, Rotnitzky A (1995). Semiparametric efficiency in multivariate regression models with missing data. Journal of the American Statistical Association.

Hahn J (1998). On the role of the propensity score in efficient semiparametric estimation of average treatment effects. Econometrica.

van der Laan MJ, Robins JM (2003). Unified Methods for Censored Longitudinal Data and Causality (Springer).

Tsiatis AA (2006). Semiparametric Theory and Missing Data (Springer).

Robins JM, Li L, Tchetgen Tchetgen ET, van der Vaart A (2008). Higher order influence functions and minimax estimation of nonlinear functionals. Probability and Statistics: Essays in Honor of David A. Freedman.

Zheng W, van der Laan (2010). Asymptotic theory for cross-validated targeted maximum likelihood estimation UC Berkeley Division of Biostatistics Working Paper Series.

Chernozhukov V, Chetverikov V, Demirer M, et al (2016). Double machine learning for treatment and causal parameters.

Examples

n <- 100; x <- matrix(rnorm(n*5),nrow=n)
a <- sample(3,n,replace=TRUE); y <- rnorm(n,mean=x[,1])

ate.res <- ate(y,a,x, sl.lib=c("SL.mean","SL.gam"))

ehkennedy/npcausal documentation built on Feb. 26, 2021, 2:43 a.m.