A Bootstrap-Based Heterogeneity Test for Meta-Analysis

Implements a bootstrap-based heterogeneity test for standardized mean differences (d), Fisher-transformed Pearson's correlations (r), and natural-logarithm-transformed odds ratio (or) in meta-analysis studies. Depending on the presence of moderators, this Monte Carlo based test can be implemented in the random- or mixed-effects model. This package uses rma() function from the R package 'metafor' to obtain parameter estimates and likelihoods, so installation of R package 'metafor' is required. This approach refers to the studies of Anscombe (1956) , Haldane (1940) , Hedges (1981) , Hedges & Olkin (1985, ISBN:978-0123363800), Silagy, Lancaster, Stead, Mant, & Fowler (2004) , Viechtbauer (2010) , and Zuckerman (1994, ISBN:978-0521432009).

boot.heterogeneity

The package boot.heterogeneity implements a Bootstrap-Based Heterogeneity Test for standardized mean differences (d), Fisher-transformed Pearson’s correlations (r), and natural-logarithm-transformed odds ratio (OR) in Meta-Analysis Studies.

Depending on the presence of moderators, this Bootstrap-Based Test can be implemented in the random or mixed-effects model. This package uses rma() function from the R package metafor to obtain parameter estimates and likelihood, so installation of R package metafor is required.

Installation

You can install the released version of boot.heterogeneity from CRAN with:

install.packages("boot.heterogeneity")

And the development version from GitHub with:

# install.packages("devtools")
# library(devtools)
devtools::install_github("gabriellajg/boot.heterogeneity")

If you have already installed this package locally, you need to override the previous version with force installation:

devtools::install_github("gabriellajg/boot.heterogeneity", force = TRUE)

Updates to this package will always be pushed to Github first. If you encounter any errors, please re-install the package from github using force installation and try again.