Reading Intervention at Age 6: Long-Term Effects of Reading Recovery in the UK on Qualifications at Age 16
Hurry, J., Fridkin, L., & Holliman, A. J. · 2022
grade Cquasi-experimentindependentmixed
Sample
293 six-year-olds, 42 London schools
Population
UK 6-year-olds with reading difficulties; followed to age-16 GCSE attainment.
Design
Three-arm (RR / standard provision in RR schools / comparison schools) non-random allocation; confounded by whole-school effects.
Key findings
A UK Reading Recovery cohort beat comparison-SCHOOL pupils at GCSE (d=0.52) but was indistinguishable from comparison pupils WITHIN the same schools (d=0.03, ns) - so most of the apparent effect is a whole-school/selection effect, not the tutoring. Illustrates how quasi-experimental 'RR works long-term' claims dissolve under the within-school counterfactual.
Genetic confound
Non-random three-arm design; the cross-school contrast is confounded by school selection.
Replication notes
WWC pooled these contrasts to 0.27 (ns), rated 'uncertain.' Points opposite to May 2023 but is weaker (grade C).
DOI / URL
10.1002/berj.3752
Effects
| Outcome | Metric | Value | Measure | Timing | Vs | Horizon | Class |
|---|---|---|---|---|---|---|---|
| GCSE qualifications: RR vs comparison schools | d | 0.52 (sig, cross-school - school-confounded) | standardized | 10yr follow-up | business-as-usual | over-2yr | attainment |
| GCSE qualifications: RR vs comparison pupils within the SAME schools | d | 0.03 (ns) | standardized | 10yr follow-up | business-as-usual | over-2yr | attainment |
Cited by
- Reading Recovery and Tier-2 1:1 early-literacy tutoringmixedconf: mediumgc: low