ASSISTments — Two Randomised Trials of an Online Maths Homework Tool with Instant Feedback for Grade 7 Students

ASSISTments is a free, non-profit online tool developed at Worcester Polytechnic Institute. Teachers assign maths homework; students receive immediate feedback on correctness and optional hints or scaffolds, and teachers see item-level reports on which concepts the class is struggling with, so they can adapt the next lesson. It is rule-based adaptive feedback, not generative AI, and is included as a rigorously evaluated benchmark for what technology-supported feedback can achieve.
Evidence: a randomised controlled trial of Grade 7 maths in Maine (roughly 43–46 schools, depending on the report) found a gain of about 0.18 standard deviations (reported as Hedges' g = 0.22 in one analysis), around half a year of extra learning. A replication in North Carolina (63 schools; disrupted by COVID-19 school closures) measured outcomes a year after implementation on the Grade 8 end-of-grade test and found an effect of +0.10. Averaged across both studies the effect is about +0.13. ASSISTments met What Works Clearinghouse standards without reservations and holds a Tier 1 (strong) rating under ESSA.
Limits: effects are modest and the North Carolina trial lost its planned end-of-year data to the pandemic, so the replication measures a delayed outcome. Results are specific to Grade 7 maths in the US and depend on teachers actually assigning and reviewing the work. The tool is not generative AI; its value for this catalogue is as a high-quality evidence benchmark.

Read the full analysis: https://www.assistments.org/research