Six design rules from the trials
- During the school day. A scheduled period, not before or after school (Nickow et al.(opens in a new tab)).
- Three or more sessions a week. Fewer sessions showed smaller effects.
- Teachers or trained aides. Paraprofessionals worked well in the Chicago trials, which kept costs down (Guryan et al.(opens in a new tab)).
- Groups of one to three. Chicago's model was two students per tutor.
- A structured program aimed at gaps. Content pitched to what the student is missing, not a repeat of today's grade-level lesson.
- Accountability and data, every week. Students own a goal and graph their weekly probe; attendance is tracked per session; a weekly data meeting applies decision rules. Progress data helps only when it changes instruction (Stecker et al., via Leiden(opens in a new tab)). The full system, with tracker and decision rules, is in Role Products.
What a 30-minute session looks like
| Minutes | Part |
|---|---|
| 0–5 | Quick review of last session from memory |
| 5–15 | Teach one skill explicitly with a worked example |
| 15–25 | Guided practice, every answer checked, immediate correction |
| 25–30 | Short independent check; tutor logs the result |
Who gets tutored. Use a screener three times a year. Start with students below benchmark in foundational skills (decoding in K–3, facts and fractions in 3–6, prerequisite algebra skills in 7–9). Review the list every six to eight weeks.
Four-week launch plan (one school)
| Week | Do | Check |
|---|---|---|
| 1 | Pick the program; schedule a daily in-school period; screen students | Is the period protected from other activities? |
| 2 | Train tutors on the program and the session structure | Can each tutor run a session as designed? |
| 3 | Launch with the highest-need groups | Attendance: are students actually getting 3+ sessions? |
| 4 | First progress check; adjust groups | Are students moving on the program's own checks? |
Troubleshooting
| Problem | Fix |
|---|---|
| Students miss sessions | Schedule it as a class period; track attendance weekly |
| Tutors reteach the class lesson | Give them the program and the student's gap data, not the class plan |
| Not enough tutors | Trained aides in groups of two or three, as in the Chicago trials |
Sources
- Nickow, Oreopoulos & Quan 2024 (opens in a new tab)
nssa.stanford.edu
- Guryan et al. 2023 (opens in a new tab)
voices.uchicago.edu
- Stecker, Fuchs & Fuchs 2005, via Leiden (opens in a new tab)
scholarlypublications.universiteitleiden.nl