vcov.rma {metafor}

Extract Various Types of Variance-Covariance Matrices from 'rma' Objects

Description

Function to extract various types of variance-covariance matrices from objects of class "rma". By default, the variance-covariance matrix of the fixed effects is returned.

Usage

## S3 method for class 'rma'
vcov(object, type="fixed", ...)

Arguments

object

an object of class "rma".

type

character string to specify the type of variance-covariance matrix to return: type="fixed" returns the variance-covariance matrix of the fixed effects (the default), type="obs" returns the marginal variance-covariance matrix of the observed effect sizes or outcomes, type="fitted" returns the variance-covariance matrix of the fitted values, type="resid" returns the variance-covariance matrix of the residuals.

...

other arguments.

Details

Note that type="obs" currently only works for object of class "rma.uni" and "rma.mv".

For objects of class "rma.uni", the marginal variance-covariance matrix of the observed effect sizes or outcomes is just a diagonal matrix with \(\hat{\tau}^2 + v_i\) along the diagonal, where \(\hat{\tau}^2\) is the estimated amount of (residual) heterogeneity (set to 0 in equal-effects models) and \(v_i\) is the sampling variance of the \(i\textrm{th}\) study.

For objects of class "rma.mv", the structure can be more complex and depends on the random effects included in the model.

Value

A matrix corresponding to the requested variance-covariance matrix.

Author(s)

Wolfgang Viechtbauer [e-mail] https://www.metafor-project.org

References

Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package. Journal of Statistical Software, 36(3), 1–48. ⁠https://doi.org/10.18637/jss.v036.i03⁠

See Also

rma.uni, rma.mh, rma.peto, rma.glmm, and rma.mv for functions to fit models for which the various types of variance-covariance matrices can be extracted.

Examples

### calculate log risk ratios and corresponding sampling variances
dat <- escalc(measure="RR", ai=tpos, bi=tneg, ci=cpos, di=cneg, data=dat.bcg)

### fit mixed-effects model with absolute latitude and publication year as moderators
res <- rma(yi, vi, mods = ~ ablat + year, data=dat)

### var-cov matrix of the fixed effects (i.e., the model coefficients)
vcov(res)

### marginal var-cov matrix of the observed log risk ratios
round(vcov(res, type="obs"), 3)

### var-cov matrix of the fitted values
round(vcov(res, type="fitted"), 3)

### var-cov matrix of the residuals
round(vcov(res, type="resid"), 3)