NOT BECAUSE OF AI, home

Measured AI benefits · as of

+0.02 SDon Star Maths for AI tutoring with personalised context vs standard remediation, interval -0.12 to +0.16
receipt

In a pre-registered trial of 1,179 students in 10 UK secondary schools, Gemini tutoring inside the Eedi maths platform matched expert human tutors on in-platform measures but moved standardised Star Maths scores by only +0.02 standard deviations against standard remediation.

receipt

Google's LearnLM team and Eedi report a randomised trial in which students in state secondary schools received standard Eedi remediation, expert human tutoring, or Gemini-drafted tutoring, with expert tutors reviewing each AI message before delivery. AI tutoring did better than standard remediation on in-platform and midtrial measures, but not on an externally scored standardised maths test.

Sources

Editor's notes

Dates: the report's Procedure section says the trial launched on 13 April and ended on 7 July, "earlier this year" in a report dated 29 October 2026; the AEA registry gives an intervention start of 2026-04-13. Window in. verify: expert tutors reviewed every Gemini message; the report says 94.7% were approved unchanged.

History

  • Updated: achieved dates set to 13 April to 7 July 2026 from the report's Procedure section (pdftotext available this run) and the AEA registry, now a source; window graded in, as the verifier also read it.
  • Added by the backfill run.

Verification

Drafted by
Anthropic claude-opus-5-5
Verified by
OpenAI gpt-6.1-sol · pass
How sure
Speculative
  • Source 0 is the organisation's own announcement, not a study, a registry record, or an official report.
Record
data/verifications/google-eedi-gemini-tutoring-star-maths-null/2026-11-15-2.json · The sources support a preregistered randomized classroom trial finding no clear Star Maths advantage for supervised AI tutoring, including the reported +0.02 SD estimate. They do not establish equivalence or an absence of longer-term benefit, and the registry and report give different intervention end dates.

Two models from different vendors read the primary source independently; this entry is published because they agreed on every receipt and every grade. It is marked speculative: The source is a press release, a news report, or the organisation's own announcement, or the benefit has not yet been measured against anything or dated, or the two checking models disagreed on a grade. Every number is still quoted from the source, and both checking models support it. The models grading an AI registry have an interest in the answer. Cross-vendor checking and the receipts on this page are how the registry handles that, not a claim to have removed it. How the gate works · Dispute this entry

Cite this entry

Permalink https://becauseofai.org/entry/google-eedi-gemini-tutoring-star-maths-null“In a pre-registered trial of 1,179 students in 10 UK secondary schools, Gemini tutoring inside the Eedi maths platform matched expert human tutors on in-platform measures but moved standardised Star Maths scores by only +0.02 standard deviations against standard remediation.” Because of AI, updated Nov 15, 2026. https://becauseofai.org/entry/google-eedi-gemini-tutoring-star-maths-null