What levers do schools have to improve math learning?

If anything moves math learning, it is what happens every day between the teacher, the students and the content. The practical question for a school or a district is therefore how to help teachers teach better, and what to expect from each way of doing so. This report compares the instructional levers for which there is accumulated causal evidence: professional learning, coaching, curriculum materials with and without support, and tutoring or small-group instruction. It compares them by their effect on learning, their cost, the scale at which that effect is achieved and how long it lasts. It follows on from the report “Everything That Matters for Learning Mathematics,” which shows that the teacher is the lever with the most weight; this report answers the next question: how to help the teacher. The program-by-program detail is in the report “What Impact Do Mathematics Programs Have on Learning?” The main conclusions are as follows:

• The lever with the largest effect per student is tutoring and small-group instruction, between +0.29 and +0.40 SD in meta-analyses and between +0.16 and +0.22 SD when measured in large-scale implementations with external tests. It is also the most expensive and the hardest to scale.
• Professional learning only moves learning when it changes what happens in the classroom. Training focused on mathematical content, without specific materials, improves what teachers know with no detectable effect on their students; individual, sustained coaching is the form of professional learning with the largest effect on practice (+0.49 SD) and a moderate effect on learning (+0.18 SD), mostly measured in reading.
• Materials work better when they come with professional learning tied to practice. The studies synthesized by Pellegrini et al. (2021) and Lynch et al. (2019) find larger effect sizes when the intervention combines curriculum materials with support and professional development (the effect size grows by +0.08 to +0.10 SD). The evidence suggests that providing good materials is not enough: helping teachers use them well increases their impact.
• The balance between cost and benefit. Intensive interventions such as tutoring or small-group work achieve, on average, large effects. Something similar happens with coaching. These interventions, however, require far more human resources and are hard to scale. Curriculum materials accompanied by training produce smaller effects per student, but they can reach practically every classroom at a relatively low additional cost.
• Effects are also judged by how long they last and how they hold up at scale. The effects of short interventions tend to fade once they stop being applied, and those of coaching and tutoring shrink by half or more as programs grow. Curriculum materials with training act every day in every classroom, for years, at a low additional cost.

Download the full report for all the details, or the visual summary to share the key idea with your team.