The R Book, 3rd Edition

17Meta‐Analysis

There is a compelling case to be made that analysts should look at the whole body of evidence rather than trying to understand individual studies in isolation. The systematic review of a body of evidence is known as meta‐analysis. The idea is to draw together all of the appropriate studies that have addressed the same question and attempted to estimate the same effect, and calculate an overall effect and an overall measure of uncertainly for it.

In an ideal world, we should be able to extract from every published study the exact question addressed, the estimated effect size, the variance of that effect, the sample size, and enough detail on the methods used to be confident that the study was comparable with the others that we have already included. This allows us to calculate a suitable overall effect size and that would be it. It sounds simple enough, but there are many pitfalls to watch out for.

We'll merely scratch the surface in this chapter, but an excellent source of information is the Cochrane Training webpage, https://training.cochrane.org/, which houses the Cochrane Handbook (Higgins et al., 2021).

17.1 Elements of a meta‐analysis

There is a lot to think about when embarking on a meta‐analysis. Broadly speaking we need to decide:

1. which studies to include in the analysis;
2. what effect is of interest and whether this is calculable from the information we have;
3. how to weight the evidence from each study;
4. what model to use for the meta‐analysis.

These ...