What researchers studied

Large-scale randomized controlled trial of a mature, comprehensive student-support program implemented in two disadvantaged urban school districts, with exploratory analyses of site- and time-specific treatment heterogeneity. 2,528 students across two urban school districts serving disadvantaged populations

What they found

  • The program's overall effect on high-school graduation was small and statistically insignificant in the full randomized sample.
  • Exploratory analyses found meaningful variation in estimated effects across sites and over time rather than one uniform program effect.
  • The pattern of heterogeneity paralleled differences in treatment timing and program implementation across settings.
  • The authors conclude that scaling a mature intervention across sites can produce results that differ from small-scale or single-site evaluations, making implementation context central to interpretation.

What the study does not prove

  • The overall randomized estimate is the strongest result; the site- and time-specific heterogeneity analyses are exploratory and should not be treated as equally definitive causal findings.
  • The study evaluates one comprehensive dropout-prevention program in two disadvantaged urban districts, so the result should not be generalized to all tutoring, coaching, mentoring, or student-support programs.
  • A statistically insignificant overall effect does not prove the program had exactly zero effect for every student or setting; it means the trial did not detect a reliable overall graduation impact in the combined sample.
  • The paper is a 2026 working paper circulated by Annenberg, NBER, IZA, and the Upjohn Institute and should not be described as a peer-reviewed journal article unless a later publication is verified.

Evidence strength: Strong experimental evidence for the overall effect; large randomized controlled trial with 2,528 students; exploratory evidence on implementation heterogeneity.

Why this matters for families

A recognizable program or comprehensive support package is not enough by itself. Families should ask how consistently the service is delivered, whether progress is measured, and what happens when a student is not responding. Implementation is part of the intervention.

Noor interpretation

How Noor translates the evidence into practice

The Noor-relevant lesson is that the name of a program is not the intervention. What families experience depends on whether the support is delivered consistently, at the intended time, with sufficient intensity and follow-through. Evidence from one site cannot be assumed to reproduce automatically when a model is expanded.

For Noor Lyra, scaling should preserve implementation quality rather than simply adding educators or students. Track whether recommended sessions actually occur, whether educators follow the instructional plan, whether families receive progress feedback, whether support begins early enough, and whether student outcomes remain consistent as the company grows. Expansion without fidelity can dilute a strong model.

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: 10.26300/zny2-0x48

Research notes

student supportdropout preventionrandomized trialimplementation fidelityhigh school graduationeducation programsscaling education

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.