Loading the library…
instruction, resources & composition
The operative claimHigh-dosage tutoring — frequent (typically >=3 sessions/week), sustained, small-group or one-to-one instruction delivered during the school day by trained tutors — is among the most reliably effective interventions for raising student achievement, with consistent positive effects across dozens of randomized field experiments; effects are largest in early grades and for lower-achieving and lower-income students.
Established that high-dosage tutoring works; the open question is scaling it to fidelity, not its efficacy.
Anchored by a meta-analysis of the experimental tutoring literature (pooled ES ~0.29 SD; Nickow et al., 2023) and convergent meta-analytic and RCT evidence (Dietrichson et al., 2017; de Ree et al., 2021). Effects are strongest for teacher/paraprofessional tutors, earlier grades, programs run during school, and higher dosage. Scaling to fidelity is the binding constraint: real-world recovery programs often deliver lower dosage than designed (Carbonari et al., 2026), and some large programs show smaller/null effects (Fryer & Howard-Noveck, 2020).
Nickow, A., Oreopoulos, P., & Quan, V (2023). The Promise of Tutoring for PreK-12 Learning: A Systematic Review and Meta-Analysis of the Experimental Evidence. American Educational Research Journal.
Meta-analysis
Tutoring produced consistently substantial positive impacts on learning (pooled ES 0.29 SD); effects were largest for teacher/paraprofessional tutors, earlier grades, programs run >=3 days/week, and tutoring delivered during school.
Bearing on this claim: Meta-analysis of tutoring RCTs: consistent positive effects (~0.29 SD); strongest for early grades, in-school, higher dosage.
doi.org/10.3102/00028312231208687Dietrichson, J., Bog, M., Filges, T., & Klint Jorgensen, A.-M (2017). Academic Interventions for Elementary and Middle School Students With Low Socioeconomic Status: A Systematic Review and Meta-Analysis. Review of Educational Research.
Meta-analysis
Tutoring (ES 0.36), feedback/progress monitoring, and cooperative learning had educationally important, robust effects for low-SES students.
Bearing on this claim: Low-SES meta-analysis: tutoring among the most effective interventions (ES 0.36).
doi.org/10.3102/0034654316687036Fryer, R. G., & Howard-Noveck, M (2020). High-Dosage Tutoring and Reading Achievement: Evidence from New York City. Journal of Labor Economics.
Experimental
130+ hours of four-on-one guided-reading tutoring improved attendance and (for Black students) ELA scores by 0.09 SD/year; mean ELA effect positive but not significant, no math effect.
Bearing on this claim: NYC RCT: high-dosage guided-reading improved attendance and (for Black students) ELA; mean effect modest.
View sourcede Ree, J., Maggioni, M. A., Paulle, B., et al (2021). Closing the income-achievement gap? Experimental evidence from high-dosage tutoring in Dutch primary education. Economics of Education Review.
Experimental
High-dosage tutoring raised math achievement by 0.28 SD after one year — about 40% of the income-achievement gap — showing the model exports across institutional settings.
Bearing on this claim: Dutch RCT: high-dosage tutoring closed ~40% of the income-achievement gap — model exports.
doi.org/10.31235/osf.io/qepc2Counter-authority (1). Studies an adversary may cite — surfaced deliberately so the grade is honest about its limits.
Carbonari, M. V., et al (2026). Impacts of Academic Recovery Interventions on Student Achievement in 2022 to 2023. Educational Evaluation and Policy Analysis.
Observational
Most programs served fewer students and lower dosage than planned; robust positive effects appeared in only 5 subject-intervention pairs — documenting how hard it is to scale intensive tutoring to fidelity.
Bearing on this claim: At-scale recovery programs: under-dosed in practice; robust effects in only 5 of many pairs — scaling is hard.
doi.org/10.3102/016237372514003774 supporting · 1 counter verified sources
Effects are largest in reading/math and early-to-middle grades; durability and transfer are less measured; and the central practical risk is dosage decay at scale, where real programs rarely match pilot intensity.
Last reviewed June 28, 2026