Sparx Maths — RAND Europe and Cambridge Find Active Use, Not Mere Access, Tracks Maths Gains

Sparx Maths is a personalised platform for Key Stage 3 and GCSE pupils (ages 11–16) with over 38,000 questions and 9,000 video tutorials. Homework is assigned by a spaced-repetition algorithm in interleaved order, with instant feedback, and teachers receive insight dashboards. Sparx recommends about one hour of homework a week (roughly 39 hours a year).
Sparx commissioned RAND Europe and the University of Cambridge to analyse data from 3,956 Year 7 and 8 pupils in 14 UK schools collected in 2019, following a study plan published before the analysis. Pupils sat the PUMA maths test. The findings were reported in January 2021.
Findings: pupils used the platform for 36.9 hours on average (range 0–256). Time on the platform was positively and significantly associated with maths outcomes; the recommended hour a week was associated with almost 20% of a predicted GCSE grade, and time actively working on problems with almost 30%. Access to Sparx alone, regardless of use, showed no detectable difference, including in a propensity-score-matching check. Effects were similar for EAL and free-school-meal pupils, so the tool neither widened nor closed attainment gaps.
Caveats: the study is observational rather than a randomised trial, was commissioned by the vendor, and the authors themselves recommend a study with a pre-established comparison group. Its main lesson is that implementation intensity, not access, drives results.

Read the full analysis: https://www.educ.cam.ac.uk/research/programmes/sparx/