## Bayesian random-effects meta-analysis

This shiny app provides a graphical user interface to the bayesmeta R package.

The bayesmeta package implements a Bayesian random-effects meta-analysis, in which several estimates are combined to a joint outcome, while allowing for a certain amount of heterogeneity between individual results. This web application provides access to the basic functionalities, some more options are available in the original R package. For a brief introduction see also the package vignette, the bayesmeta documentation, or the following article:

- C. Röver. Bayesian random-effects meta-analysis using the bayesmeta R package. Journal of Statistical Software,
**93**(6):1–51, 2020.

In the panel on the left you may specify prior settings, and in the "data" tab above you can input data. Once you are done, hit the "compute" button on the left. You may then inspect the results in the other tabs above.

This app is distributed under the GNU general public license; you can download the source code here.

Christian Röver, 2022.

## Data to be analyzed

You need to enter the data to be analyzed below. First specify the number of studies (k), then fill in estimates y_{i} and associated standard errors σ_{i} in the "y" and "sigma" columns. You can also provide labels for each entry.

Alternatively, you can also use one of the provided example data sets to get started. Data may also be imported from (or exported to) a 3-column CSV file. For an example, see e.g. this data set.

## Analysis output summary

## Forest plot

## Joint posterior density

## Marginal posterior density (effect μ)

## Marginal posterior density (heterogeneity τ)

## Credible interval calculator

## R code

The R code to reproduce the main computations is shown below.

You only need to fill in the "y", "sigma" and (optionally) "labels" arguments.

For more detailed help, please consult the package vignette or the bayesmeta documentation.