The 60-second card
The finding
Curriculum is not neutral. In a large randomized trial, the math program a school was assigned changed how much its students learned.
Do this
Choose materials with a clear sequence, teacher-led instruction of new content, worked examples, and built-in review. Judge them against this library, not only against standards coverage.
Best evidence
A federal randomized trial assigned one of four math curricula to 110 elementary schools in 12 districts. By the end of 2nd grade, students using Math Expressions, Saxon, and Scott Foresman–Addison Wesley/enVision had all outperformed students using Investigations (Mathematica(opens in a new tab); IES(opens in a new tab)).
110 schools
in a randomized trial of four math curricula
The 5-minute brief
Why it matters. Curriculum decisions affect every classroom at once and last for years, which makes them one of the biggest levers a district has. The trial is rare: curricula are almost never compared head to head with random assignment.
The classroom differences are telling. Student-centered instruction and peer collaboration were highest in Investigations classrooms; teacher-directed instruction was highest in Saxon classrooms. Saxon teachers also spent about an hour more per week on math, largely because of its daily routines (IES(opens in a new tab)). In second grade, more frequent requests for multiple strategies and solutions went with lower achievement (IES evaluation summary(opens in a new tab)).
What to look for in materials
| Feature | What good looks like |
|---|---|
| Sequence | Skills build in a clear order; fact fluency and fractions are not delayed |
| New content | Teacher explains and models before students practice independently |
| Examples | Worked examples paired with practice problems |
| Review | Daily review and mixed practice from earlier units, built in |
| Knowledge (ELA) | Connected units in science, history, and the arts, not stand-alone skill passages |
| Checks | Frequent quick checks with guidance on what to reteach |
Common mistakes
- Choosing materials by standards alignment alone. Aligned materials can still teach in ways the evidence doesn't support.
- Adopting strong materials and then cutting the parts teachers find repetitive, such as daily review.
What a leader looks for in 5 minutes
- Is the adopted curriculum actually being used, as designed?
- Are the review and practice parts taught, or skipped for time?
What we don't know yet
Only a handful of curricula have ever been compared in randomized trials, and a six-state comparison study found little difference in math gains by textbook at typical levels of use and training (Blazar et al. 2020(opens in a new tab)). The four tested here were early-grade math programs in schools serving disadvantaged students, and newer editions may differ from what was tested.
Go deeper
- Researcher
- Roberto Agodini (study director)
- Translator
- Anna Stokke and Jonathan Regino on choosing math resources (Chalk & Talk episode 73)
Also called: high-quality instructional materials, HQIM, curriculum adoption