Call us anytime

Last updated: November 11, 2026
A good math tutor finds the exact step where your child's understanding breaks, then works with them one-on-one or in a very small group, several times a week, until it holds. That combination is one of the best-proven ways to raise math achievement. A review of tutoring experiments published by the National Bureau of Economic Research found "consistent and substantial" gains, and in University of Chicago trials, high school students who received daily math tutoring gained the equivalent of one to two and a half extra years of math in a single year. The first step is to name the goal: catching up, keeping up or getting ahead. Below: how to tell whether your child needs a tutor, what makes tutoring work, the main types of tutors and what they cost, how to choose one, and what to do when your child is ready for more than the next chapter.
The tutors who get the fastest results all start the same way, and it is not with homework. Here is what they look for in the first session. Show me how to find it →
Math builds on itself. A gap in fractions in grade 5 becomes a problem in algebra in grade 8, so small gaps are worth closing early. On the 2024 National Assessment of Educational Progress (NAEP), the national test run by the US Department of Education:
| Grade | At or above NAEP Proficient | At NAEP Advanced | Below NAEP Basic |
|---|---|---|---|
| Grade 4 | 39% | 9% | 24% |
| Grade 8 | 28% | 8% | 39% |
NAEP notes that its Proficient level is a demanding standard and does not mean the same thing as "on grade level" on a state test. Still, the pattern is clear: math gets harder for more students in middle school, and only about 1 eighth grader in 12 reaches Advanced.
Signs a tutor could help:
The research on tutoring is unusually strong. In a 2020 review of experimental studies, Andre Nickow, Philip Oreopoulos and Vincent Quan found that tutoring programs produced "consistent and substantial positive impacts," with an average effect of 0.37 standard deviations. Effects were larger for tutoring by teachers and trained paraprofessionals, and math tutoring tended to show larger effects in the later grades.
The Chicago trials, run by the University of Chicago Education Lab with the nonprofit Saga Education, tested daily 50-minute sessions with two students per tutor during the school day. Published in the American Economic Review, the results showed math test scores rising by 0.18 to 0.40 standard deviations, better grades in math and other courses, and effects that lasted into later years.
Stanford's SCALE Initiative summarizes what the strongest programs share:
| Feature | What the research supports |
|---|---|
| Frequency | Three or more sessions a week |
| Time | About 90 minutes a week in total |
| Length | At least 10 weeks |
| Group size | No more than four students per tutor; one-on-one gives the strongest gains in online tutoring |
| Consistency | The same tutor each time, which improves attendance, motivation and learning |
| Results | Learning gains of three to more than 15 months |
You do not need a school program to use these findings. They tell you what to ask any tutor for: regular sessions, enough of them, and the same person each week.
Three sessions a week with the same tutor can add months of learning. The families who see it happen ask four questions before they hire anyone. Send me the insider guide for parents →
| Option | What your child gets | Best for |
|---|---|---|
| The teacher's own help | Extra help sessions, office hours, the teacher's view of the gap | Every child, first |
| Free nonprofit help online | Khan Academy's free lessons and practice; free live peer tutoring on Schoolhouse.world | Practice, review and quick unsticking |
| Peer or student tutor | An older student who recently took the same course | Keeping up in a specific class |
| Private one-on-one tutor | A plan built for one child, with direct feedback | A specific, stubborn gap, or moving ahead |
| Learning center or small group | A set program and a regular schedule | Children who do well with structure and peers |
| AI tutor | Step-by-step help at any hour | Extra practice between sessions |
Khan Academy is a nonprofit, and it says its content library and learning platform "will always remain 100% free." Schoolhouse.world, a nonprofit founded in 2020 by Khan Academy's founder Sal Khan, runs free, live tutoring sessions over Zoom, led by trained student volunteers, including homework help in math. Both are excellent places to start. For how to use AI tools well, see our guide to AI tutoring.
One point parents raise often: a tutor who makes grade 7 math click is not automatically the right person for calculus. Match the tutor to the course.
No official source publishes what families pay for math tutoring. The closest official data is what tutors earn. In May 2025, the US Bureau of Labor Statistics counted 175,070 employed tutors, a category that includes math, other subjects and test prep:
| Hourly pay for employed tutors, May 2025 | Amount |
|---|---|
| 25th percentile | $16.83 |
| Median | $20.84 |
| Mean | $23.10 |
| 75th percentile | $26.38 |
| 90th percentile | $36.53 |
Two cautions. These figures cover employees and exclude self-employed tutors, and a family's hourly rate is not the same as a tutor's wage. Rates also rise with the level of math, the tutor's experience, one-on-one versus group format and where you live.
So is $30 or $40 an hour a lot? Both sit above what most employed tutors earn: $30 is above the 75th percentile and $40 is above the 90th. That does not make either rate unfair for an experienced private tutor who teaches advanced math. The better question is the one in our SAT tutoring guide: what will your child be able to do after these hours that they cannot do now?
The hourly rate is the least useful number in the decision. The number that matters most can be measured after three sessions. Show me how to find it →
Talk with at least two tutors. Ask each one:
Then watch your child after the first few sessions. A good match shows up as less dread at homework time, and a child who can explain a step they could not explain before.
Tutoring is not only for catching up. Many families use a tutor to preview a course before it starts, or to move faster than the class. If your child is clearly ahead, also ask the school about subject acceleration or curriculum compacting; our guide on how to challenge a gifted child explains both.
Competitions give a strong math student peers and a goal. MATHCOUNTS is the classic middle school path, the AMC 10 and AMC 12 are the first step toward the AIME and the USA Mathematical Olympiads, and students who love code can try USACO.
Some children want something else again: not a faster version of school math, but a real question. Mathematicians spend their careers on problems in game theory, combinatorics, network science and the mathematics of music, and a curious teenager can start on one. This is the step past tutoring. A tutor helps your child master the material the class covers. A research mentor helps them go somewhere the class never will.
That is the work learnSTEM does: one-on-one research with a PhD mentor from a top-20 university, on a question your child chooses, carried from first idea to finished paper. As one learnSTEM student wrote, "I got to learn many new notions in math that are much more complex compared to my standard math classes in school. Yet, I managed to understand each piece of information clearly." For a younger child, our guide on how to get into research in middle school shows how it starts.
Imagine your child at 14, explaining a question in combinatorics that no worksheet ever asked. See how bright math students make that leap. Show me the experiments →
No official source reports an average price. The Bureau of Labor Statistics reports that employed tutors earned a median of $20.84 an hour in May 2025, with the top 10% above $36.53. Private rates vary with the tutor's experience, the level of math, the format and your location.
It is above what 90% of employed tutors earn, according to BLS data for May 2025. For an experienced private tutor in advanced math, it can still be fair value. Judge it by the plan and the results.
Stanford's review finds the evidence is stronger for in-person tutoring, but online tutoring has also shown meaningful positive effects in high-quality studies, with one-on-one online sessions producing the strongest gains. Choose the format your child will attend every week.
The strongest programs meet three or more times a week, about 90 minutes in total, for at least 10 weeks. For a small gap, once or twice a week may be enough; ask the tutor for a plan with a clear end point.
For practice and quick help, it can be very useful. Research so far finds the strongest results when a person stays involved. Our AI tutoring guide explains how to set one up well.
All bright students reach a point where school stops challenging them. learnSTEM provides that challenge: the space to explore a field of their choice, guided by a world-class PhD mentor from a top-20 university, completely one-on-one and on your own schedule. Students finish with a passion for a new subject, a research project they are proud of, and a clear sense of what they want to pursue next.
See if your child qualifies for the STEM Explorers Program → · See if your child qualifies for the STEM Research Program → · Questions about eligibility? Contact us →
"It has only strengthened my love for learning about the world and given me a paper that I am proud to say I've made." — learnSTEM student
Early Admission – November 15, 2026
Regular Admission – December 13, 2026
Late Admission – January 31, 2027
The Program runs four times per year: Spring, Summer, Fall, and Winter.
