Bayesian Evidence Synthesis: New Meta-Analytic Procedures

Eric-Jan Wagenmakers, Raoul Grasman, Quentin F. Gronau, and Felix Schönbrodt developed a suite of meta-analytic techniques for Bayesian evidence synthesis, addressing a series of challenges that currently constrain classical meta-analytic procedures. These techniques include (1) an application of bridge sampling to obtain Bayes factors for random-effects meta-analysis; (2) the computation of Bayes factors for fixed-effect versus random-effects meta-analysis; (3) proposal of an informed prior on study heterogeneity based on a comprehensive literature search; (4) model-averaged evidence across fixed-effect and random-effects meta-analyses, thereby accounting for model-uncertainty; and (5) a proposal for a running power analysis in the field of meta-analysis.
These techniques allow for the (1) quantification of evidence, both for and against the absence of an effect; (2) monitoring of evidence as new studies accumulate over time; (3) graceful and principled “model-averaged” combination between fixed-effect and random-effects meta-analysis; and (4) principled planning of a new study in order maximize the probability that it will lead to a worthwhile gain in knowledge.