In September 2024, a widely-shared analysis in Education Week reported that recent large-scale trials of federally-funded tutoring produced effects roughly one-third the size of the deluxe research programs schools had been told to emulate. Within weeks, the framing had shifted across superintendent LinkedIn feeds: high-dosage tutoring, once the consensus response to pandemic learning loss, was being written off as another intervention that didn't scale. For charter directors and special programs coordinators who spent 2022 and 2023 building tutoring lines into their budgets, the reversal creates a real problem. Boards and authorizers now ask whether the investment was worth it. The honest answer requires reading the studies more carefully than the headlines. What the research actually shows is not that tutoring failed. It shows that most programs labeled "high-dosage" never delivered the dose.
The narrative pivot traces to a set of large randomized trials evaluating ESSER-funded tutoring at scale, including work published by the National Student Support Accelerator at Stanford and evaluations commissioned by Accelerate, the national nonprofit incubated and launched by America Achieves. These studies found effect sizes in the range of 0.03 to 0.10 standard deviations, real but small, and well below what the original evidence base predicted. Coverage in Education Week, The 74, and Chalkbeat has treated these results as a verdict on the intervention itself. Susanna Loeb, who directs the NSSA at Stanford, has been more careful in her public statements, pointing out that the trials measured what happened when tutoring was scaled through district procurement, not what happened when tutoring was delivered as designed.
If you sit on the intervention side of a charter or district budget, you are now defending line items in a different rhetorical environment than the one you built them in. Board members read the same headlines your families read. Authorizers ask harder questions during renewal. Federal program coordinators face pressure to redirect Title I and Title III funds toward interventions that look, on the surface, less contested. The temptation is to disinvest, to cut the tutoring line and shift dollars into a curriculum adoption or a coaching contract that carries less narrative risk. In our experience working with charter intervention programs, that reflex is a mistake, but only if you can articulate the distinction between what the studies measured and what your program actually does. A program that delivers 40 sessions to 90 percent of enrolled students is not the same intervention as a program that delivered 17 sessions to 76 percent of enrolled students, even if both used the same purchase order line and the same vendor label.
The foundational evidence base for high-dosage tutoring comes from the Nickow, Oreopoulos, and Quan meta-analysis published as NBER Working Paper 27476 in July 2020, which pooled 96 randomized evaluations and reported an average effect size of roughly 0.37 standard deviations. That number is what schools were told to expect. It is also what got detached from its methodological context. As Monica Bhatt, coauthor of the University of Chicago Education Lab research on scaled tutoring, has put it, prior research had focused on "deluxe" programs that cost multiple thousands of dollars per student to ensure students received regular tutoring. Small tutor-to-student ratios, trained tutors, tightly aligned curriculum, and per-pupil costs that ran into the several thousands of dollars precisely because they included the operational infrastructure required to make sure students actually showed up.
The recent scaled evaluations tell a different story about implementation. The NSSA and Accelerate analyses of ESSER-era programs report that the median student in the studied cohorts received approximately 17 tutoring sessions across a school year, against program designs that specified 40 to 60. Participation rates hovered near 76 percent. Session length varied. Tutor turnover was high. Denver schools, for instance, contracted with two external companies to conduct tutoring; one struggled to find qualified tutors, the other conducted tutoring virtually with easier staffing but frequent missed sessions, and many students in the district didn't get the recommended hours of tutoring. When researchers modeled effect sizes against dose actually delivered, the small pooled effect resolved into a familiar pattern. Students who received closer to the full designed dose showed effects more consistent with the original evidence base. Students who received a fraction of the dose showed a fraction of the effect. The intervention did not fail. The delivery system around it did.
Real high-dosage tutoring, as defined in the NSSA Tutoring Quality Standards, requires one-on-one or small-group sessions with no more than four students per tutor, use of high-quality materials that align with classroom content, at least three sessions per week of at least 30 minutes each, sessions held during school hours, students meeting with the same tutor each session, and professionally trained tutors who receive ongoing support and coaching. The reason so many ESSER-funded programs missed these standards is not that vendors were dishonest. It is that the standards are operationally expensive to meet, and the procurement structures schools used, including hourly billing, opt-in enrollment, and pull-out scheduling contested with electives, did not incentivize providers to protect them.
When was the last time you audited the median dose actually delivered to your Tier 3 cohort, as opposed to the dose your program design specifies?
A+ Tutoring's iLEAD Tier 3 Math case study offers one data point on what happens when the implementation architecture is protected. In the 2024-25 school year, A+'s Tier 3 cohort at iLEAD Exploration showed 75 percent of Math Tier 3 students (9 of 12) reaching MAP growth benchmarks, 87.5 percent of ELA Tier 3 students (7 of 8) hitting the same standard, and 80 percent of the combined Tier 3 cohort (16 of 20) reaching benchmarks at 3 to 6 times national MAP Growth norms. The variables that produced those numbers are the ones the failed studies were missing: participation floors written into the partnership MOU, session counts audited monthly against a coordinator's dashboard, pairings held stable across semesters, and tutor onboarding designed for continuity rather than turnover. The tutoring itself is not novel. The delivery discipline is.
A+ Tutoring holds the NSSA Tutoring Program Design Badge (2024-2026), awarded by the National Student Support Accelerator at Stanford University after an evidence-based review that signifies the quality of our program design and our alignment to Tutoring Quality Standards, as assessed by researchers and practitioners. The Badge speaks to how our program is designed, not to the outcomes above, which stand on the case-study data.
If your Tier 3 cohort were named individually, would the same students appear in next year's cohort because the intervention never reached them?
Regardless of whether you ever talk to A+, five steps are worth taking this quarter. First, ask your current tutoring provider for the median session count actually delivered to enrolled students in the last completed semester, not the design target. Second, ask for the participation rate, meaning the percentage of enrolled students who received at least 75 percent of the designed sessions. Third, audit your intervention scheduling to see whether pull-out time is being contested by electives, specials, or unrelated pull-outs; in our experience, unresolved schedule conflicts are among the largest practical drivers of dose shortfall. Fourth, compare tutor-student pairing stability across the year. In our experience running intervention cohorts, pairing continuity is a bigger driver of student engagement than tutor credentialing on paper, which is why NSSA's own quality standards call for students meeting with the same tutor each session. Fifth, if you sit on the federal programs side of the budget, pull the Title I Part A and Title III supplement-not-supplant documentation for your current tutoring contract and confirm that the deliverables described in your program narrative match the dose your provider is actually delivering. In our experience, mismatch between program-narrative deliverables and actual dose delivered is one of the most common findings we see when partners walk us through their monitoring documentation.
A+ Tutoring is a California K-12 virtual intervention provider working with charter schools and district partners to deliver high-dosage tutoring aligned to NSSA Tutoring Quality Standards. A+ partner schools have shown 75 percent of Math Tier 3 students reaching growth benchmarks, 87.5 percent in ELA Tier 3, and 80 percent in the combined Tier 3 cohort, at 3 to 6 times national MAP Growth benchmarks.
Walk Your Tier 3 Dose Data With Danielle