chapter  22
WithBlakeley B. McShane, Ulf Böckenholt
Pages 14

Meta-analysis is a well-established statistical technique that synthesizes two or more studies of a common phenomenon. Because multiple studies provide more information about the common phenomenon than any single one of them, meta-analysis can offer a number of benefits. For example, insofar as the studies measure the common phenomenon with some degree of error, a meta-analysis, which pools the results from the studies via a weighted average, will yield an estimate that is on average more accurate than that of any individual study. In addition, the uncertainty in the meta-analytic estimate will typically be smaller than the uncertainty in the estimate of any individual study. Further, meta-analysis allows for the investigation of differences among studies, for example by quantifying the impact of study-level covariates or the degree of between-study variation.