In colleenchan/metamediate: Estimate Pooled Total Effects in Meta-Analysis

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)

metamediate

The goal of metamediate is to estimate a pooled total effect in meta-analysis.

Installation

You can install the most updated version of metamediate as follows:

library(devtools)
devtools::install_github("colleenchan/metamediate")

Example

library(metamediate)
# Simulate a meta-analysis
S <- 30
ST <- 10
SB <- 10
SD <- 10
set.seed(1)
beta.T <- rnorm(ST + SB, mean = 0.3, sd = 0.05)
set.seed(1)
beta.D <- rnorm(SD + SB, mean = 0.2, sd = 0.05)
set.seed(1)
var.beta.T <- rnorm(ST + SB, mean = 0.05, sd = 0.01)
set.seed(1)
var.beta.D <- rnorm(SD + SB, mean = 0.05, sd = 0.01)
df <- data.frame(beta.T = c(beta.T, rep(NA, SD)),
                  beta.D = c(rep(NA, ST), beta.D),
                  var.beta.T = c(var.beta.T, rep(NA, SD)),
                  var.beta.D = c(rep(NA, ST), var.beta.D))

# Estimate summary mediation proportion and pooled effect
res <- metamed(te = df$beta.T,
                te_lb = df$beta.T - qnorm(0.975) * sqrt(df$var.beta.T),
                de = df$beta.D,
                de_lb = df$beta.D - qnorm(0.975) * sqrt(df$var.beta.D),
                author = paste("Study", seq(1:S)),
                rr = "identity")