Frozen normal random-effects benchmarks. Optional public-data runners re-estimate them from prepared inputs; OCRS uses the public-use descriptive complete-case school contrasts.
| application | J | pooled_effect | pooled_ci_lower | pooled_ci_upper | tau2_REML | I2_percent | prediction_lower | prediction_upper |
|---|---|---|---|---|---|---|---|---|
| star | 79 | 0.1936935 | 0.0965344 | 0.290853 | 2.15257e-02 | 11.18856850 | -0.1098354 | 0.497222 |
| bangert_drowns | 48 | 0.2219296 | 0.1317036 | 0.312156 | 4.99411e-02 | 58.37398254 | -0.2252699 | 0.669129 |
| ocrs | 24 | 0.0373988 | -0.0773139 | 0.152111 | 3.33088e-07 | 0.00040174 | -0.0773194 | 0.152117 |