If there is one intervention that the research community agrees on, it is tutoring. Not all tutoring. Not the kind where a college student sits with a high schooler for 30 minutes once a week and works through homework. The kind that works is high-dosage, frequent, and integrated into the school day. The effect sizes are among the largest in all of education research.
High-dosage tutoring, defined as three or more sessions per week of 30+ minutes each, produces large and consistent academic gains in both math and reading. The critical design elements are frequency (at least three sessions per week), integration into the school day (not after school), small group size (one-to-three students per tutor), and consistent tutor-student relationships. Programs that meet these criteria consistently produce large gains. Programs that compromise on frequency or group size see dramatically reduced effects.
What makes tutoring high-dosage
Frequency matters most
The single biggest factor is how often students receive tutoring. Three sessions per week is the minimum threshold for meaningful impact. Daily tutoring produces even larger effects. Once-a-week tutoring produces minimal results regardless of quality.
This is not intuitive. Many districts operate tutoring programs with one or two sessions per week and wonder why outcomes do not improve. The research is clear: below three sessions per week, the dosage is insufficient.
Integration into the school day
After-school tutoring programs consistently underperform tutoring delivered during the school day. The reason is simple: attendance. After-school programs face chronic attendance issues, and the students who need tutoring most are often the ones who cannot stay after school due to transportation, family obligations, or after-school jobs.
During-the-school-day tutoring eliminates the attendance barrier. It also signals that tutoring is a core part of the educational program, not a remedial add-on.
Small groups
The most effective programs use a one-to-one or one-to-three tutor-to-student ratio. Once group size exceeds four students, the individualization that makes tutoring effective begins to erode. A tutor with eight students is not tutoring. They are teaching a small class, which is a different intervention with different outcomes.
Consistent relationships
Students who work with the same tutor over time show larger gains than students who rotate between tutors. The relationship between tutor and student builds trust, allows the tutor to understand the student's specific gaps, and creates accountability that motivates consistent effort.
Who should do the tutoring
Certified teachers produce the largest effects
This is expensive but unambiguous. Programs staffed by certified teachers produce the largest academic gains. Districts that use existing staff (repurposing coaching positions, using interventionists, or building tutoring into teacher schedules) see the strongest results.
Trained paraprofessionals are effective
Well-trained paraprofessionals produce significant gains, typically 60-80% of the effect size produced by certified teachers. The key word is trained. A paraprofessional with structured materials, ongoing coaching, and clear protocols can be highly effective. A paraprofessional without these supports will struggle.
College students and volunteers vary widely
Volunteer and college student tutors produce inconsistent results. Some programs using college students show meaningful gains. Others show almost none. The difference is usually the quality of training, supervision, and materials provided. If you use volunteer tutors, invest heavily in the support structure around them.
Practical design considerations
Where does the time come from?
This is the hardest question. If tutoring happens during the school day, students are missing something else. The most common approaches: replacing a study hall or elective period, using an intervention block built into the master schedule, or extending the school day by 30 minutes specifically for tutoring.
The intervention block approach is the most sustainable. Build a 30-minute daily block into the master schedule. Students who need tutoring receive it. Students who do not participate in enrichment activities. This requires master schedule redesign but produces the most consistent implementation.
How do you identify students?
Use multiple data points: diagnostic assessments, classroom teacher recommendations, and prior year achievement data. Reassess every six to eight weeks and adjust student groupings. Tutoring should not be permanent for any student. The goal is accelerated growth until the gap closes, then transition back to core instruction.
What to measure
- Tutoring attendance rate (if students are not showing up, nothing else matters)
- Dosage delivered (actual sessions per student per week vs. intended)
- Student growth on diagnostic assessments (measured every 6-8 weeks)
- Tutor quality and consistency (observation and student survey data)
- Comparison to non-tutored peers (are tutored students growing faster than similar non-tutored students?)
Common mistakes
- Scheduling tutoring after school and wondering why attendance is low. If you want consistent attendance, put tutoring during the school day.
- Using group sizes of 6-8 and calling it tutoring. That is small group instruction, which is valuable but produces smaller effects than true tutoring.
- Operating once-a-week programs. Below three sessions per week, the research shows minimal impact. Do not spend resources on insufficient dosage.
- Rotating tutors frequently. Consistency of the tutor-student relationship drives results. Minimize tutor rotation.
If you only do one thing this week: Count the number of tutoring sessions each of your tutored students receives per week. If it is less than three, you are below the research threshold. Increasing frequency to three-plus sessions will produce more improvement than any other change you could make to the program.