A demonstration and evaluation of the use of cross-classified random-effects models for meta-analysis
information for practice,
Abstract It is common for the primary studies in meta-analyses to report multiple effect sizes, generating dependence among them.