idata

A randomly generated dataset containing 2000 cases 7 columns with no missing values.
The intermediate confounders are assumed to be independent of each other.

datasets

We implement causal decomposition analysis using methods proposed by Park, Lee, and Qin (2022) and Park, Kang, and Lee (2023), which provide researchers with multiple-mediator imputation, single-mediator imputation, and product-of-coefficients regression approaches to estimate the initial disparity, disparity reduction, and disparity remaining (<doi:10.1177/00491241211067516>; <doi:10.1177/00811750231183711>). We also implement sensitivity analysis for causal decomposition using R-squared values as sensitivity parameters (Park, Kang, Lee, and Ma, 2023 <doi:10.1515/jci-2022-0031>). Finally, we include individualized causal decomposition and sensitivity analyses proposed by Park, Kang, and Lee (2025+) <doi:10.48550/arXiv.2506.19010>.

Suyeon Kang

causal.decomp

Causal Decomposition Analysis

Soojin Park

Karen Xu

idata function

A data frame containing the following variables. The data are provided only for
explanatory purposes.<dl>
<dt>Y:</dt>
<dd>A continuous outcome variable.</dd><dt>R:</dt>
<dd>A binary group indicator with a value of 0 (reference) and 1 (comparison).</dd><dt>M:</dt>
<dd>A binary risk factor with a value of 0 (not treated/received) and 1(treated/received).</dd><dt>X1:</dt>
<dd>First continuous intermediate confounder.</dd><dt>X2:</dt>
<dd>Second continuous intermediate confounder.</dd><dt>X3:</dt>
<dd>Third continuous intermediate confounder.</dd><dt>C:</dt>
<dd>A continuous baseline covariate.</dd>
</dl>

Format

Synthetic Data for illustrating optimal treatment regimes and individualized effects — idata

A data frame containing the following variables. The data are provided only for
explanatory purposes.<dl>
<dt>Y:</dt>
<dd>A continuous outcome variable.</dd>

<dt>R:</dt>
<dd>A binary group indicator with a value of 0 (reference) and 1 (comparison).</dd>

<dt>M:</dt>
<dd>A binary risk factor with a value of 0 (not treated/received) and 1(treated/received).</dd>

<dt>X1:</dt>
<dd>First continuous intermediate confounder.</dd>

<dt>X2:</dt>
<dd>Second continuous intermediate confounder.</dd>

<dt>X3:</dt>
<dd>Third continuous intermediate confounder.</dd>

<dt>C:</dt>
<dd>A continuous baseline covariate.</dd>


</dl>

idata: Synthetic Data for illustrating optimal treatment regimes and individualized effects

Description

Usage

Arguments

Format

Details

References