A Meta-Analysis on the Optimal Cumulative Dosage of Early Phonemic Awareness Instruction
Erbeli, F., Rice, M., Xu, Y., Bishop, M. E., & Goodrich, J. M. · 2024
grade Cmeta-analysisindependentreplicated
Sample
16 studies / 35 effect sizes (613 tx / 542 control), mean age 5.20
Population
Preschool through grade 1; includes students at-risk for reading disability.
Design
Nonlinear dose-response meta modeling cumulative instructional hours against effect size, separately for speech-only vs letter-integrated PA.
Key findings
Answers the dose question: speech-only PA peaks at ~10.2 hours (d=0.74) then DECLINES - extended pure-PA drilling is counterproductive - while letter-integrated instruction keeps paying off past 16 hours. Reinforces 'brief PA, integrated with letters.'
Genetic confound
Experimental/quasi-experimental base.
Replication notes
Converges with Rice 2022 (moderators ns) and Rehfeld 2022 (no dose-response) that brief PA suffices.
DOI / URL
10.1080/10888438.2024.2309386
Effects
| Outcome | Metric | Value | Measure | Timing | Vs | Horizon | Class |
|---|---|---|---|---|---|---|---|
| Speech-only (basic) PA, peak effect | d | 0.74 at ~10.2 cumulative hours; declines beyond (inverted-U) | mixed | end-of-treatment | unclear | end-of-treatment | domain-skill |
| Letter-integrated PA | shape | effects keep rising past 16 hours (no early plateau) | mixed | end-of-treatment | unclear | end-of-treatment | domain-skill |
Cited by
- Phonemic awareness training (and whether it needs letters)moderate supportconf: mediumgc: low