meta_default: Defaults for Model Averaging in Meta-Analysis

Description

Wrapper with default prior for Bayesian meta-analysis based on a literature review. Currently, the same default is used in all cases.

Usage

meta_default(y, SE, labels = NULL, field = "psychology", effect = "ttest",
  ...)

Arguments

mean in each study

standard error in each study

labels

optional: character values with study labels

field

either"psychology" or "medicine" (uses partial matching, so "p" and ""m" are sufficient)

effect

the type of effect size: either means ("ttest"), log-odds ratios ("logOR") or (Fisher's z-transformed) correlations ("corr") (also uses partial matching)

...

further arguments passed to meta_bma

Details

Default prior distributions can be plotted using plot_default.

For field = "psychology", the following defaults are used:

effect = "ttest": Half-normal with SD=0.3 on mean effect and half-Cauchy with scale=.5 on standard deviation of effects.
effect = "logOR": Half-normal with SD=0.3 on mean effect and half-Cauchy with scale=.5 on standard deviation of effects.
effect = "corr": Half-normal with SD=0.3 on mean effect and half-Cauchy with scale=.5 on standard deviation of effects.

For field = "medicine", the following defaults are used:

effect = "ttest": Half-normal with SD=0.3 on mean effect and half-Cauchy with scale=.5 on standard deviation of effects.
effect = "logOR": Half-normal with SD=0.3 on mean effect and half-Cauchy with scale=.5 on standard deviation of effects.
effect = "corr": Half-normal with SD=0.3 on mean effect and half-Cauchy with scale=.5 on standard deviation of effects.

References

Gronau, Q. F., Erp, S. V., Heck, D. W., Cesario, J., Jonas, K. J., & Wagenmakers, E.-J. (2017). A Bayesian model-averaged meta-analysis of the power pose effect with informed and default priors: the case of felt power. Comprehensive Results in Social Psychology, 2(1), 123-138. https://doi.org/10.1080/23743603.2017.1326760

Examples

Run this code

# NOT RUN {
data(towels)
d1 <- meta_default(towels$logOR, towels$SE, towels$study,
                   field = "psych", effect = "logOR",
                   sample = 5000)
d1
plot_forest(d1)
# }