Photo provided by DepositPhotos
Only 28% of 8th graders in the United States are proficient in math. That statistic, drawn from the 2024 National Assessment of Educational Progress, reflects a learning gap that has persisted despite more than $190 billion in federal pandemic recovery spending. For parents watching their child fall behind, the numbers raise an uncomfortable question: is the tutoring they're paying for actually working?
A wave of peer-reviewed research from Johns Hopkins, Stanford, and Brown universities is now offering some answers, and the findings should factor into any family's decision about tutoring.
People are also reading…
The Proficiency Problem Is Not Going Away
Post-pandemic learning loss has proven stubbornly resistant to broad interventions. Federal relief funds underwrote everything from extended school days to group tutoring programs, yet national math scores remain below 2019 levels. The gap is sharpest among middle school students, who have the least time to recover before high school math sequences begin.
This context matters when parents compare tutoring options. Choosing a program based on brand recognition or proximity is reasonable. Choosing based on what the research says produces measurable academic gains is better.
What the Science Says About Format
Not all tutoring works the same way. A 2025 report from the Alliance for Learning Innovation, drawing on Johns Hopkins research, found that high-dosage one-on-one tutoring produces three to four months of additional learning, outperforming every other school-based academic intervention, including summer school and reduced class sizes.
The format of delivery also matters less than many parents assume. New quasi-experimental studies show that well-designed virtual one-on-one sessions achieve effect sizes nearly identical to in-person tutoring (.21 versus .24). Poorly structured group models, by contrast, show effect sizes as low as .05, a marginal return on a significant investment of time and money.
Stanford and Brown research adds a finer point. Middle school students in individual tutoring sessions gained approximately eight more percentile points than peers placed in groups of three. That gap compounds over an academic year.
Center-Based Programs: What Parents Should Know
Franchise tutoring centers like Mathnasium and Kumon are among the most recognized names in the market, and for good reason. They offer structured curricula, physical locations, and recognizable accountability frameworks. For some students, particularly those who respond well to routine and in-person engagement, these programs represent a meaningful option.
The research does raise questions worth asking before enrolling. Center-based models typically involve group instruction or semi-supervised independent work rather than true one-on-one sessions. Based on the effect size data above, parents should ask directly: how much individualized attention will my child receive per session?
Cost and scheduling are practical considerations as well. Center-based programs often require multi-month commitments and consistent in-person attendance, which can create friction for families with variable schedules.
For parents weighing their options between Mathnasium and Kumon, independent comparisons have grown more detailed in recent years, examining cost, scheduling, and academic outcomes across all three formats.
Online Tutoring: The Evidence Is Catching Up
The perception that online tutoring is a lesser substitute for in-person instruction has not held up under scrutiny. The near-identical effect sizes for virtual and in-person one-on-one sessions suggest the medium is secondary to the model.
School administrators appear to agree. A 2023–24 National Center for Education Statistics School Pulse Panel found that 90% of schools providing high-dosage tutoring rated it as at least moderately effective, with 52% calling it 'very' or 'extremely' effective.
The clearest finding across the research is this: dosage and individualization predict outcomes far more reliably than location.
Making a Decision That Reflects the Data
Parents comparing tutoring programs are navigating a market shaped more by marketing than by published outcomes data. The research now available gives families a sharper lens.
The questions worth asking any provider are straightforward: How many students share attention during each session? What is the ratio of instructor time to student work time? How are sessions adapted when a student plateaus?
Programs that answer those questions clearly, and align with the one-on-one, high-dosage model the research supports, are the ones most likely to produce the gains families are looking for.
