ks_data: Generate Kang & Schafer (2007) Scenarios

In dewittpe/cbal-v1: Covariate Balancing Weights using Generalized Projections of Bregman Distances

Description

The ks_data function generates a dataset according to the experimental scenarios that appear in Kang and Schafer (2007). These scenarios are useful for evauluating different covariate balancing methods under treatment and/or outcome model mispecification.

Usage

ks_data(tau, n, sig2, rho, y_scen = c("a", "b"), z_scen = c("a", "b"))

Arguments

tau	the causal effect to be estimated.
n	the sample size.
sig2	the marginal error variance of the potential outcomes. We assume heteroscedasticity.
rho	the correlation coefficient between the potential outcomes.
y_scen	defaults to c("a", "b")
z_scen	defaults to c("a", "b")

References

Kang, J. D., & Schafer, J. L. (2007). Demystifying double robustness: A comparison of alternative strategies for estimating a population mean from incomplete data. Statistical science, 22(4), 523-539.

dewittpe/cbal-v1 documentation built on July 2, 2020, 6:24 p.m.