← Back to registry

Synthetic recovery

Effect-Size Benchmarks for Education

Kraft (2021) · Education Policy

T1T2T3T4T5

All five gates passed

Power analysis produces sample sizes consistent with Kraft benchmarks

T1 RealityT2 BootstrapT3 InvariantsT4 LedgerT5 Two calculators
N validSample size for d = 0.20 at 80% power
d = 0.20 is ‘large’Published value
N/A95% confidence interval
1,000Simulated students
25Items

Description

Validation that the engine’s power analysis produces sample sizes consistent with Kraft’s benchmarks. Kraft argues d = 0.20 is a large effect in field conditions.

Methodology

Simulated 1,000 learners across 25 items and ran the engine’s internal power analysis to determine the minimum sample size needed to detect d = 0.20 at 80% power.

Reference

Kraft, M. A. (2020). Interpreting effect sizes of education interventions. Educational Researcher, 49(4), 241–253.

This is a synthetic recovery, not a replication of the original study. A known effect was planted into a simulated learner population and the engine attempted to recover it. Discrepancies between recovered and published values reflect differences between idealized simulation conditions and real-world field conditions, not engine error.