Study duration had to be at least 12 weeks
Outcome measures had to be assessments of the mathematics being taught in all classes. Almost all are standardized tests or state assessments.
The review placed particular emphasis on studies in which schools, teachers, or students were assigned at random to experimental or control groups.