get_im04: Estimation of multiplier for confidence interval for ATE as...

In matteobonvini/sensitivitypuc: Sensitivity Analysis via the Proportion of Unmeasured Confounding (PUC)

Description

get_im04 computes c_alpha as in Imbens and Manski (2004).

Usage

get_im04(dat, n, alpha)

Arguments

dat	vector containing estimate of lower bound, estimate of upper bound, estimate of standard deviation of lb, estimate of standard deviation of ub in this order.
n	sample size.
alpha	desired significance level (capped at pnorm(5)).

Value

a scalar equal to the multiplier used to construct the confidence interval for partially identified ATE.

References

Imbens, G. W., & Manski, C. F. (2004). Confidence intervals for partially identified parameters. Econometrica, 72(6), 1845-1857.

Examples

n <- 100
dat <- c(-0.4, 0.3, 0.2, 0.1)
get_im04(dat, n, 0.05) # large identification interval width
dat2 <- c(-0.4, -0.395, 0.2, 0.1)
get_im04(dat2, n, 0.05) # small identification interval width

matteobonvini/sensitivitypuc documentation built on Dec. 9, 2020, 2:24 a.m.