Is a good tutor worth it? What the research says

How much does tutoring actually help?

Nickow, Oreopoulos, and Quan pooled randomized trials of preK-12 tutoring and found an average effect of 0.37 standard deviations (NBER Working Paper 27476, July 2020). The J-PAL summary of the same review puts that in plain terms: a student at the 50th percentile moves to nearly the 66th. J-PAL calls anything above 0.3 a large effect.

Small trials tend to report larger effects, so the 2021 EdResearch for Action brief by Robinson, Kraft, Loeb, and Schueler also reports the number for scale: studies of 15 larger programs, each serving 500 to 7,000 students, averaged 0.25 standard deviations. That is smaller, and still a real gain.

The same brief translates the meta-analysis into time: roughly 3 to 15 extra months of learning, depending on the program and grade. That months figure is the brief's translation, not a separate study.

How often does a tutor need to come?

The EdResearch brief describes the pattern behind the strong results: three or more sessions a week, about 30 to 60 minutes each, for at least 10 weeks. Many effective programs ran a full school year. Elementary students sometimes did better with 20-minute sessions five days a week.

Once a week has little evidence behind it. The brief cites Roland Fryer's 2016 review of 196 field experiments (NBER Working Paper 22130), which defined high-dosage tutoring as more than three days a week, or at least 50 hours over 36 weeks. In that review, high-dosage tutoring was 20 times more effective than low-dosage tutoring in math and 15 times more effective in reading.

The J-PAL summary adds detail on the third session. Moving from one or two sessions a week to three helped at every grade. A fourth or fifth day only helped preschool through first grade. For grades two through five, three sessions averaged 0.37 standard deviations and four or five averaged 0.28.

Timing matters too. School-day programs averaged 0.40 standard deviations against 0.21 for after-school programs, and the brief puts the same gap at roughly twice as large. J-PAL notes the comparison only covers paraprofessional and nonprofessional programs, since teacher tutoring in the sample all ran during the school day.

Does it matter who the tutor is?

Yes. The pooled effects by tutor type in the J-PAL summary: teachers 0.50 standard deviations across 17 studies, paraprofessionals 0.40 across 47, parents 0.23 across 11, nonprofessional volunteers 0.21 across 24. Paraprofessionals here means school staff, education undergraduates, and service fellows who got training and a structured curriculum.

One caveat the review flags itself: five of the 17 teacher studies evaluated Reading Recovery, a program with intensive training and a fixed curriculum, and that pulls the teacher average up. The lesson is not the job title. It is training, structure, and support.

The EdResearch brief says tutors can effectively teach up to three or four students at a time, which is how a family can split the cost.

I am Sam. I run this site. Tutors were a game changer for me in school. They are how I kept up with advanced classes. I did not need someone to sit in the room while I guessed. I needed someone who could teach the next idea before the class moved on.

What does a tutor cost?

Care.com's rate guide (updated September 22, 2025) lists 2025 averages of $24.40 an hour for elementary tutors, $24.37 for middle school, and $24.22 for high school, with test prep near $100 and tutoring centers at $50 to $75. Its cost pages show a national average of $25.59 an hour as of September 2026. Wyzant's own guide (December 2024) puts the typical rate on its marketplace at $35 to $60 an hour.

Now price the schedule the research supports. Three one-hour sessions a week for 10 weeks is 30 hours. At $25 an hour that is $750. At $50 an hour it is $1,500. Pair your child with one classmate at 2:1 and each family pays half.

Compare the plan most families buy: one hour a week for 10 weeks. That is $500 at $50 an hour, cheaper on paper, and it is the dose the studies found little evidence for.

What 10 weeks at 3 sessions looks like

Copy this and change the subject. It assumes a 4th grader behind in fractions and 45-minute sessions on Monday, Wednesday, and Friday. For a kindergartner or first grader, use 20 minutes, five days a week.

• Week 1: the tutor works through the last unit test and current homework with your child, names the two or three gaps in writing, and sends you the plan. No written plan, ask for one.
• Weeks 2 to 4: reteach gap one (say, equivalent fractions) at the start of every session, then 15 minutes on what the class is doing this week so homework stops piling up. Between sessions: one 10-minute printable, then stop.
• Week 5: check. The tutor re-gives a quiz from the gap unit. Passing means gap two starts. Not passing means the plan changes, not the child.
• Weeks 6 to 9: gap two, plus the current unit. Ask the tutor for one sentence after each week: what got better, what is next.
• Week 10: re-test on the original unit and read the latest class grade. Decide to continue, taper to once a week for maintenance, or stop. Do not let it drift.

When is a tutor worth it, and when is it not?

Ask the school first. Many districts run high-dosage programs during the school day, the setting with the best results, at no cost to you. If you hire privately, buy the ingredients the studies used before you buy credentials.

PrintableSpark worksheets and games are for the days between sessions. They give your child independent reps without turning you into the teacher. They do not replace the person who spots the gap.

Put your real calendar into the estimator below. It tells you whether the plan you are pricing sits inside the range the studies used, or whether you are about to pay for a once-a-week plan with a three-a-week price tag.

• Worth it: the same tutor every session, 3 or more sessions a week, 30 to 60 minutes, at least 10 weeks, a written plan tied to the current class unit, and a check at week 5.
• Worth it: a 2:1 or 3:1 group of similar-level kids with a trained tutor, which cuts the price without cutting the evidence.
• Not worth it: one session a week that finishes homework and nothing else.
• Not worth it: a new tutor every few weeks. Spotting the gap and building trust both take time.
• Not worth it: paying for a plan nobody re-tests. If no one can tell you what improved by week 10, stop.