rmeta: Rank-based Meta-analysis

In YandaLang/RankBasedMeta: Rank-based Mata-analysis

Rank-based Meta-analysis

Description

Rank-based Meta-analysis

Usage

rmeta(data, effect = "fixed", method = "MMNP", fullsample = TRUE)

Arguments

data	input data
effect	fixed effects or random effects
method	Meta-analysis methods, default MMNP
fullsample	data type. full sample or summary statistics for limited information
alpha	significance level, default alpha=0.05
iteration	number of iterations for bootstrap, only used for random effect bootstrap methods

Note

format for full sample: data[,1] <- study1_treatment; data[,2] <- study1_control; data[,3] <- study2_treatment; data[,4] <- study2_control;...

format for summary statistics: data[1,] <- c(study1_treatmentlocation, study1_treatmentscale, study1_treatmentsamplesize); data[2,] <- c(study1_controllocation,study1_controlscale,study1_controlsamplesize); data[3,] <- c(study2_treatmentlocation,study2_treatmentscale,study2_treatmentsamplesize);...

format for summary statistics - two sample hodges-lehmann: data[1,] <- c(study1_shift,study1_scale,study1_treatmentlocation,study1_controllocation,study1_treatmentsamplesize,study1_controlsamplesize); data[2,] <- c(study2_shift,study2_scale,study2_treatmentlocation,study2_controllocation,study2_treatmentsamplesize,study2_controlsamplesize);...

Author(s)

Yanda Lang, Joseph McKean, Omer Ozturk

Examples

sample1_treat <- rnorm(25,0,1)
sample1_con <- rnorm(25,0,1)
sample2_treat <- rnorm(25,1,1)
sample2_con <- rnorm(25,1,1)
dat <- data.frame(sample1_treat,sample1_con,sample2_treat,sample2_con)
rmeta(dat, effect = "fixed", method = "MMNP", fullsample = TRUE)

YandaLang/RankBasedMeta documentation built on July 20, 2023, 12:44 p.m.