Effectiveness of Reading-Strategy Interventions in Whole Classrooms: A Meta-Analysis
Okkinga, M., van Steensel, R., van Gelderen, A. J. S., et al. · 2018
grade Cmeta-analysisindependentmixed
Sample
52 studies / 125 effect sizes
Population
Whole-classroom reading-strategy interventions, school-age.
Design
Meta focused on ecologically valid whole-class delivery; separates standardized from researcher-developed outcomes and researcher- vs teacher-delivery.
Key findings
Under realistic whole-class delivery, strategy instruction moves standardized comprehension only d=0.19 (vs 0.43 on researcher-made tests), and effects shrink further when teachers rather than researchers deliver. The honest at-scale estimate is ~0-0.19.
Genetic confound
Experimental base; effects inflate when the researcher (not the teacher) delivers.
Replication notes
Quantifies the whole-class/teacher-delivered deflation the older 0.32 anchor missed.
Effects
| Outcome | Metric | Value | Measure | Timing | Vs | Horizon | Class |
|---|---|---|---|---|---|---|---|
| Comprehension (standardized measures) | d | 0.186 | standardized | end-of-treatment | business-as-usual | end-of-treatment | domain-skill |
| Comprehension (researcher-developed measures) | d | 0.431 | researcher-designed | end-of-treatment | business-as-usual | end-of-treatment | domain-skill |
| Strategic-ability outcomes | d | 0.786 | researcher-designed | end-of-treatment | business-as-usual | end-of-treatment | near-transfer |
Cited by
- Reading-comprehension strategy instruction (reciprocal teaching, summarizing, questioning)mixedconf: mediumgc: low