The
metafor package consists of a collection of functions for conducting meta-analyses in
R. Fixed- and random-effects models (with and without moderators) can be fitted via the general linear (mixed-effects) model. For 2x2 table data, the Mantel-Haenszel and Peto's method are also implemented. Functions for complex syntheses (such as network meta-analyses, multivariate meta-analyses, and meta-analyses with correlated outcomes) will be implemented in the future.
The package grew out of the mima function (see below), which I now consider obsolete. My goal is to provide a single package for conducting all sorts of meta-analyses in R.
Please post any questions about the package to the
R-help mailing list. That way, others can benefit from the information and answers provided. Make sure you first read and follow the
posting guide. To make sure that my mail filter catches your post, include the package name (i.e., metafor) in your post.
You can obtain and install the package directly within R by typing: install.packages("metafor") (assuming an internet connection and appropriate access rights on the computer). Alternatively, you can download the source files or a Windows binary provided here and install them manually. See the
R Installation and Administration manual for more details on how to install packages this way. Note that the version of the source file and binary provided here may be slightly newer than the version available on
CRAN and from within R.
metafor_0.5-7.tar.gz -- source package
metafor_0.5-7.zip -- Windows binary
metafor.pdf -- manual (in pdf format)
news.txt -- NEWS file for the package (lists the changes/updates to the package)
The figures below illustrate some of the types of plots that can be drawn with the metafor packge.
plot_forest1.png -- a forest plot with some additional annotations added
plot_forest2.png -- a forest plot with predicted odds ratios added
plot_funnel.png -- a contour-enhanced funnel plot
plot_radial.png -- a radial plot
plot_qqnorm.png -- a Q-Q normal plot
plot_influence.png -- a plot of various case diagnostics
Current Version: 0.5-7
