QTE: IPW estimator for the Quantile Treatment Effect

In pedrohcgs/IPS: IPS: Covariate Distribution Balance via Integrated Propensity Scores

Description

IPW estimator for the Quantile Treatment Effect

Usage

QTE(
  y,
  d,
  x,
  ps,
  beta.lin.rep,
  tau = 0.5,
  bw = "nrd0",
  trim = FALSE,
  trim.at = NULL,
  whs = NULL
)

Arguments

y	An n x 1 vector of outcome of interest.
d	An n x 1 vector of binary treatment adoption indicators.
x	An n x k matrix of covariates used in the propensity score estimation
ps	An n x 1 vector of fitted propensity scores.
beta.lin.rep	An n x k matrix of estimates of the asymptotic linear representaion of the propensity score parameters (used to compute std. errors).
tau	An l x 1 vector of quantile to compute the QTE at. If NULL, then we set tau = 0.5.
bw	Bandwidth choice to compute densities (used to compute std. errors). Options are "ucv","nrd", "nrd0", "bcv", "SJ" - see bw.nrd for additional details. Default choice is "nrd0".
trim	Logical argument to whether one should trim propensity scores. Deafault is FALSE.
trim.at	Only used if trim=TRUE. If a scalar, trim all propensity score below trim.at and above 1 - trim.at. If a 2 x 1 vector, trim all propensity scores below trim.at[1] and all propensity scores above trim.at[2]. If NULL, trim.at is set to 1e-10.
whs	An optional n x 1 vector of weights to be used. If NULL, then every observation has the same weights.

Value

A list containing the following components:

qte	The estimated QTE
qte.se	Estimated (pointwise) std. error of the QTE.
qte.inf	Estimated influence function of QTE estimator.
tau	The evaluation points of QTE.

References

Sant'Anna, Pedro H. C, Song, Xiaojun, and Xu, Qi (2019), Covariate Distribution Balance via Propensity Scores, Working Paper <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3258551>.

pedrohcgs/IPS documentation built on Dec. 22, 2021, 7:39 a.m.