rrma: Robust random-effect meta-regression with dependent effect...

In seananderson/robustmeta: Robust random-effect meta-regression with dependent effect sizes

Description

rrma returns robust variance estimates from random effect meta-regression with dependent effect-size estimates.

Usage

  rrma(formula, data, study_id, var_eff_size, rho)

Arguments

formula	The meta-regression formula of the form y ~ x1 + x2..., where y is the effect size. An intercept only meta-analysis can be performed with y ~ 1. The intercept can be excluded with y ~ -1 + x1 + x2 ...
data	The input data frame.
study_id	The study IDs. Can be in any form (character, numeric, factor). Will be converted to factor and then numeric form internally.
var_eff_size	The variance on each effect size.
rho	The assumed correlation between effect sizes of the same study including correlation induced by the random effects. Hedges et al. (2010) suggest conducting a sensitivity analysis on rho.

Details

This is an implementation based on the robust.se function from Hedges et al. (2010).

Value

A list with the robust covariance matrix, the Qe statistic, the tau_sq estimate (variance between studies), a data frame of the robust coefficient estimates and standard errors, and a number of diagnostics and input values.

Author(s)

Main body of code is from the appendix of Hedges et al. (2010) and written by the paper authors. Code adapted to an R package with a formula interface and various convenience functions by Sean Anderson and Jarrett Byrnes.

References

Hedges, L.V., Tipton, E. & Johnson, M.C. (2010). Robust variance estimation in meta-regression with dependent effect_size estimates. Res. Synth. Method., 1, 39-65.

Examples

data(broad)
(m <- rrma(formula = lnorReg ~ d18OresidualMean.cent, data =
broad, study_id = study.ID, var_eff = vlnorReg, rho = 0.5))

seananderson/robustmeta documentation built on May 29, 2019, 4:25 p.m.