README.md

In achambaz/mediation.test: More Powerful Tests of the Composite Null Hypothesis Arising in Mediation Analysis

mediation.test

More Powerful Tests of the Composite Null Hypothesis Arising in Mediation Analysis

The mediation.test R package implements more powerful tests of the composite null hypothesis arising in mediation analysis. It is developed by A. Chambaz (MAP5, UMR CNRS 8145, Université Paris Cité) \& Caleb Miles (Department of Biostatistics, Columbia University Mailman School of Public Health).

Introduction

The mediation_test() function of the mediation.test package tests the composite null hypothesis "δx * δy = 0" against its alternative "δx * δy ≠ 0".

> library("mediation.test")
>
> example(mediation_test_minimax)
> example(BH_mediation_test_minimax)
>
> example(compute_map_rejection_probs)
> example(mediation_test_Bayes)

Citation

To cite the package, see

> citation("mediation.test")
> toBibtex(citation("mediation.test"))

Installation

Both a stable version and a development version are available via GitHub and can be installed in R as:

devtools::install_github("achambaz/mediation.test", ref = "main")

or

devtools::install_github("achambaz/mediation.test", ref = "develop")

achambaz/mediation.test documentation built on Oct. 20, 2024, 9:25 a.m.