Key findings

Primary maths

Educator's summary

As a PDF document

As web pages:
- Key findings
Top-rated programmes
Programmes with limited evidence
Other programmes
Review methods

Full report

As a PDF document (113 pages)

What works in teaching maths?

Report summary as a PDF document

Instructional Process Strategies (IP) (Changing the way the teacher teaches, eg co-operative learning)

The highest-quality studies and strongest positive effects were found for instructional process programmes such as co-operative learning, classroom management and motivation programmes, and small-group tutoring programmes. Median effect size across 36 studies: +0.33.

Information and Communication Technology (ICT) (Programmes based on ICT)

Our review showed that most studies of ICT find small positive achievement outcomes. However, the outcomes are very mixed, and the highest-quality studies find few positive effects. Also, most qualifying studies evaluated programmes that are no longer available; there are few studies of current versions of ICT. Median effect size across 38 studies: +0.19. (An effect size of less than 0.20 is weak).

Mathematics Curricula (MC)

Our review found little evidence that it matters which textbook is used, at least for pupil outcomes on standardised tests. Studies of innovative curricula supported by the US National Science Foundation, such as Everyday Mathematics and Math Trailblazers, found small differences in maths achievement in comparison to control groups. Similarly, Saxon Math and traditional maths texts had little evidence of effectiveness. Median effect size across 13 studies: +0.10. (An effect size of less than 0.20 is weak).

Overall, 87 studies met the inclusion criteria, of which 36 used random assignment to treatments. These included 13 studies of mathematics curricula (2 randomised), 38 studies of ICT (15 randomised), and 36 studies of instructional process programmes (20 randomised).