Fixed Effect vs Random Effects Meta-Analyses

There are two methods of meta-analysis: Fixed-effect and Random effects. In a Fixed-effect meta-analysis, we assume that all included studies are testing the same ‘true’ effect size, for example, the correlation between drinking intentions and drinking behaviour, and that studies with larger samples provide a more precise estimate of this true effect size. This means assigning these studies greater weight in a meta-analysis. In contrast, a Random effects meta-analysis assumes that the ‘true’ effect size is likely to vary between studies due to differences in how studies were conducted, where they were conducted, et cetera. The goal of this chapter is to introduce the ideas of fixed-effect meta-analysis, contrast these with random effects meta-analysis, before outlining reasons to prefer random effects meta-analysis.