Table 3.
Overview of the estimators for the between‐study variance.
| Estimator | Abbreviation | Iterative/Non‐iterative | Positive/Non‐negative |
|---|---|---|---|
| Method of moments estimators | |||
| DerSimonian and Laird | DL | Non‐iterative | Non‐negative |
| Positive DerSimonian and Laird | DLp | Non‐iterative | Positive |
| Two‐step DerSimonian and Laird | DL2 | Non‐iterative | Non‐negative |
| Hedges and Olkin | HO | Non‐iterative | Non‐negative |
| Two‐step Hedges and Olkin | HO2 | Non‐iterative | Non‐negative |
| Paule and Mandel | PM | Iterative | Non‐negative |
| Hartung and Makambi | HM | Non‐iterative | Positive |
| Hunter and Schmidt | HS | Non‐iterative | Non‐negative |
| Maximum likelihood estimators | |||
| Maximum likelihood | ML | Iterative | Non‐negative |
| Restricted maximum likelihood | REML | Iterative | Non‐negative |
| Approximate restricted maximum likelihood | AREML | Iterative | Non‐negative |
| Model error variance estimator | |||
| Sidik and Jonkman | SJ | Non‐iterative | Positive |
| Bayes estimators | |||
| Rukhin Bayes | RB | Iterative | Non‐negative |
| Positive Rukhin Bayes | RBp | Iterative | Positive |
| Full Bayes | FB | Iterative | Non‐negative |
| Bayes Modal | BM | Iterative | Positive |
| Bootstrap estimator | |||
| Non‐parametric bootstrap DerSimonian and Laird | DLb | Iterative | Non‐negative |