What matters most here
| Rank | Practice | Grade |
|---|---|---|
| 1 | Mixed practice sets across units (principle 8) | Evidence grade: Proven in classrooms Proven; randomized trial in 54 seventh-grade classes, 61% vs 38% a month later (IES(opens in a new tab)) |
| 2 | Low-stakes quizzing, including older material (principles 6, 7) | Evidence grade: Proven in classrooms Proven (Yang et al.(opens in a new tab)) |
| 3 | Worked examples, faded as students improve (principle 3) | Evidence grade: Strong Strong |
| 4 | Fraction and proportion gaps closed before algebra (principle 11) | Evidence grade: Strong Strong in principle; trials were in grade 4 |
A lesson in this room
Warm-up: three problems from three earlier units. Teacher works a slope example step by step. Students do a twin problem with the last step blank, then a full one, checked on whiteboards. Homework: three slope problems mixed with proportions, circle area, and inequalities.
Leader look-fors
- Practice sets that include earlier units, mixed in
- A worked example on the board for new problem types
- Warm-ups answered from memory
Stop doing
Homework sets of 20 problems on today's lesson only. Most textbooks are built this way: under 10% of problems in six common texts were mixed (Rohrer et al.(opens in a new tab)).
Where the evidence runs out
Effects on state tests are still being tested in an independent replication due by 2027 (IES(opens in a new tab)).
Principles
Sources
- Rohrer et al. 2020 (IES) (opens in a new tab)
ies.ed.gov
- Yang et al. 2021, classroom quizzing (opens in a new tab)
scholars.hkbu.edu.hk
- Rohrer et al. 2020, textbooks (opens in a new tab)
link.springer.com
- WestEd replication of interleaved math practice (opens in a new tab)
ies.ed.gov