In an earlier post about sandwich standard errors for multi-variate meta-analysis, I mentioned that Beth Tipton has recently proposed small-sample corrections for the covariance estimators and t-tests, based on the bias-reduced linearization approach of McCaffrey, Bell, and Botts (2001). You can find her forthcoming paper on the adjustments here. My understanding is that these small-sample corrections are important because the uncorrected sandwich estimators can lead to under-statement of uncertainty and inflated type I error rates when a given meta-regression coefficient is estimated from only a small or moderately sized sample of independent studies (or clusters of studies). Moreover, it can be difficult to determine exactly when you have a large enough sample to trust the uncorrected sandwiches.

I wanted to try out these small-sample corrected sandwich estimators for a meta-analyses project that I’m working on. Beth and one of her students have written an R package called robumeta that implements the sandwich covariance estimator and small-sample corrections as described in her paper. However, for my project I want to use the metafor package, which doesn’t provide these methods. I’ve therefore created a set of functions that implement the sandwich covariance estimators and small-sample corrections for models estimated using the `rma.mv`

function in `metafor`

. Here is the complete code. Sorry, there’s no further documentation at the moment (beyond the rest of this post).

Back to top