Bias-adjusted posterior-predictive benchmarks for STEM moderator effect sizes (0.18, 0.38, 0.64 at the 25th, 50th, 75th percentiles) fall below both Cohen's conventional thresholds and raw empirical quartiles¶

Claim · All claims
Evidence · 1 study (2 entries) · quant-synthesis r3 · q3
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i? Empirical quartiles of observed moderator effect magnitudes were 0.29, 0.53, and 0.77, while publication-bias-adjusted posterior-predictive quartiles shifted downward to 0.18, 0.38, and 0.64. → Bicakci 2026 q3 i? Pronounced jumps in the raw distribution around Cohen's conventional thresholds disappeared after publication-bias adjustment, suggesting these thresholds may function as implicit reporting incentives. → Bicakci 2026 (2)

Evidence¶

Bicakci 2026¶

Bicakci, M., Heller, F., Stoeger, H., & Ziegler, A. (2026). From RAMSing to Rigor: Improving Moderator Analysis in STEM Education Meta-Research. Educational Psychology Review. https://doi.org/10.1007/s10648-026-10172-1

q3 · i? · quant-synthesis · r3

Benchmarking synthesis of 1,786 absolute g values from 89 meta-analyses (distribution right-skewed, mean 0.59, median 0.53) contrasted with 10,000 posterior-predictive draws from RoBMA-PSMA (Monte Carlo SE = 0.003).

"The observed data yielded quartiles of 0.29, 0.53, and 0.77, whereas the fully adjusted posterior- predictive distribution shifted downward to 0.18, 0.38, and 0.64."

Bicakci 2026 (2)¶

Bicakci, M., Heller, F., Stoeger, H., & Ziegler, A. (2026). From RAMSing to Rigor: Improving Moderator Analysis in STEM Education Meta-Research. Educational Psychology Review. https://doi.org/10.1007/s10648-026-10172-1

q3 · i? · quant-synthesis · r3

Histogram comparison of the raw observed distribution (Fig. 5) against the bias-adjusted posterior-predictive distribution (Fig. 6); the authors read the raw discontinuities as possible secondary RAMSing driven by magnitude thresholds.

"Notably, the pronounced jumps around Cohen’s (1988) conventional thresholds in the raw distri bution (see Fig. 5) disappeared after publication-bias adjustment (Fig. 6)."

Discussion¶

Related Claims¶

• Publication-bias adjustment via limit meta-analysis lowers the overall STEM benchmarks to 0.14, 0.32, and 0.59 — related
• Benchmarks are sensitive to precision and publication bias, but subgroup ordering remains stable and one-per-primary-study quartiles are broadly similar to raw values — related
• The bias-adjusted mean moderator effect in STEM education meta-analyses is g = 0.27 with substantial heterogeneity, and the published record shows overwhelming evidence of publication-bias-type selection — related
• Under bias-adjusted planning assumptions, approximately 14 independent moderator effect sizes per moderator level are needed for 80% power, and about 60.2% of corpus effect sizes fell below this threshold — related