Comparative role of various methods of estimating between study variance for meta-analysis using random effect method

Abstract

Methods of synthesizing the effect size is guided by observed heterogeneity. As a convention, fixed effect method (FEM) is used for low/no heterogeneity. However, random effect method (REM) is used for substantial heterogeneity. But, synthesized (i.e., pooled) effect size under REM also relies on the method used to estimate between study variance along with within study variance. There are various methods to assess between study variance to be used under REM. Accordingly, present study compared existing methods of estimating between study variance on the basis of coverage probability and precision in order to find preferred method of assessing between study variance. Data from a systematic review and meta-analysis for various outcomes involving varying extent of sample size and heterogeneity was used. Hunter and Schmidt method and DerSimonnian & Laird method were found as preferred methods to estimate between study variance.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 9 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.