varmeta: MAP approach for variances

In tobiasmuetze/varmap: Tools Related to MAP Priors for Variances

MAP approach for variances

Description

Performs a random effects meta-analysis for variances.

Usage

varmeta(sample_var, degf, mu_mean, n.iter = 10000, ...)

Arguments

sample_var	numeric vector with the sample variances from historical clinical trials
degf	numeric vector with the degrees of freedom of the sample variances
mu_mean	mean of the normal prior for the mean
n.iter	number of iterations to monitor in the R function coda.samples
...	Additional arguments for the R functions coda.samples and jags.model

Details

The function varmeta performs a random effect meta-analysis for variance. The model is as proposed by Schmidli et al., i.e. the log-variance are modeled using a normal hierarchical model with normal prior of the mean and a half-normal prior for the (between-trial) standard deviation.

Value

An mcmc.list object with the posterior predictive for the variance (map_variance), the posterior for the mean of the log variance (mu), and the between-trial standard deviation of the log-variance (tau)

References

Schmidli H, Neuenschwander B, Friede T. Meta-analytic-predictive use of historical variance data for the design and analysis of clinical trials. Computational Statistics & Data Analysis (2016).

tobiasmuetze/varmap documentation built on Dec. 8, 2022, 2:16 p.m.