Analysis of categorical moderators in mixed‐effects meta‐analysis: Consequences of using pooled versus separate estimates of the residual between‐studies variances | Publicación