What researchers studied

Expanded meta-analysis of 263 randomized controlled tutoring trials, with analyses that progressively align the evidence with large-scale programs and independent standardized outcomes. 263 randomized controlled trials of tutoring programs

What they found

  • The full sample produced a pooled achievement effect of about 0.40 standard deviations.
  • Studies with stronger equivalence to large-scale programs targeting independent standardized outcomes produced pooled effects of about 0.16 to 0.22 standard deviations.
  • The decline was strongly associated with program scale.
  • A bundled package of commonly recommended tutoring design features partially reduced the attenuation observed at scale.

What the study does not prove

  • Meta-analytic moderator patterns do not prove that each individual design feature independently causes better outcomes.
  • Large-scale programs differ in population, staffing, subject, delivery model, measurement, and implementation context.
  • The 0.16–0.22 SD estimates are policy-relevant pooled expectations for a target class of programs, not a promised effect for an individual student or provider.
  • The study focuses substantially on standardized achievement outcomes and does not capture every possible benefit of tutoring.

Evidence strength: Peer-reviewed meta-analysis of 263 randomized controlled trials.

Why this matters for families

Large tutoring effects from small studies should not be treated as automatic expectations at scale. The useful lesson is to ask whether the features that made support effective are still present when a program grows.

Noor interpretation

How Noor translates the evidence into practice

Growth should not be treated as a reason to relax instructional standards. As tutoring expands, educator continuity, dosage, ratios, instructional materials, diagnosis, and quality assurance become more—not less—important to monitor.

Audit completed dosage, educator continuity, student-to-educator ratios, use of diagnostic evidence, session documentation quality, and independent outcomes as enrollment grows. Track whether quality indicators deteriorate with scale rather than assuming the model transfers automatically.

Read the original source

Noor links to the original or authoritative source so families can distinguish the evidence itself from our interpretation.

Open original source →

DOI: 10.3102/00346543261446660

Research notes

tutoring at scalemeta-analysisrandomized trialstutoring qualityimplementation fidelitystudent achievement

No single study determines a student's plan. Noor uses research as one input alongside the learner's goals, observed performance, academic context, and response to instruction.