Skip to contents

This function calculates statistics related to the test of treatment effect consistency across groups.

Usage

consist.test(T, tau, Y, ngates = 5, nsim = 10000)

Arguments

T

A vector of the unit-level binary treatment receipt variable for each sample.

tau

A vector of the unit-level continuous score. Conditional Average Treatment Effect is one possible measure.

Y

A vector of the outcome variable of interest for each sample.

ngates

The number of groups to separate the data into. The groups are determined by tau. Default is 5.

nsim

Number of Monte Carlo simulations used to simulate the null distributions. Default is 10000.

Value

A list that contains the following items:

stat

The estimated statistic for the test of consistency

pval

The p-value of the null hypothesis (that the treatment effects are consistent)

Details

The details of the methods for this design are given in Imai and Li (2022).

References

Imai and Li (2022). “Statistical Inference for Heterogeneous Treatment Effects Discovered by Generic Machine Learning in Randomized Experiments”,

Author

Michael Lingzhi Li, Operations Research Center, Massachusetts Institute of Technology mlli@mit.edu, http://mlli.mit.edu;

Examples

T = c(1,0,1,0,1,0,1,0)
tau = c(0,0.1,0.2,0.3,0.4,0.5,0.6,0.7)
Y = c(4,5,0,2,4,1,-4,3)
consisttestlist <- consist.test(T,tau,Y,ngates=5)
consisttestlist$stat
#> [1] 0.01949332
consisttestlist$pval
#> [1] 0.686