Generalized confidence intervals for effect quantiles — ci_lower_quantile

Computes confidence intervals for effect quantiles that can be generalized to a larger population using the hypergeometric correction approach.

Usage

ci_lower_quantile_generalize(
  Z,
  Y,
  N,
  k_vec = ceiling(c(0.6, 0.7, 0.8, 0.9) * N),
  alpha = 0.05,
  gamma = 0.5,
  ndraw = 10^4,
  treat.method.list = list(name = "Stephenson", s = 6),
  control.method.list = list(name = "Stephenson", s = 6),
  score = NULL,
  stat.null = NULL,
  nperm = 10^4,
  Z.perm = NULL,
  set = "all",
  alpha.ratio.treat = 0.5,
  tol = 10^(-3),
  simul = TRUE
)

Arguments

Z: An n-dimensional binary treatment assignment vector (1 = treated, 0 = control).
Y: An n-dimensional observed outcome vector.
N: Population size.
k_vec: Vector of quantile ranks of interest.
alpha: Significance level.
gamma: Proportion of alpha for hypergeometric correction.
ndraw: Number of Monte Carlo draws.
treat.method.list: Method specification for treated units.
control.method.list: Method specification for control units.
score: Optional pre-computed score vector.
stat.null: Optional pre-computed null distribution.
nperm: Number of permutations for null distribution.
Z.perm: Optional permutation matrix.
set: Set of quantiles: "treat", "control", or "all".
alpha.ratio.treat: For set="all", proportion of alpha allocated to treated.
tol: Tolerance for root-finding.
simul: Logical; if TRUE, compute simultaneous intervals.

Value

A data frame with columns k, lower, and upper.