Claims
Each claim is one testable statement that circulates about AI in education, assessed separately for secondary and university learners. Under it sits every finding we have linked to it, sorted by whether it supports the claim, contradicts it, or doesn't really test it.
Statuses are the collection's own nine-label scale, defined in full on the About page. They are claim-level judgments, not GRADE ratings. A claim still in review is not listed.
01Theme
Unassisted performance after assisted practice
- Unrestricted use of a general-purpose LLM chatbot during practice improves secondary students' subsequent performance on assessments taken without AI.Not supported · Secondary students · Evidence stable · 3 linked findings (0 for, 1 against)
- Unrestricted use of a general-purpose LLM chatbot during practice improves university students' subsequent performance on assessments taken without AI.Insufficient evidence · University students · Evidence unclear · 5 linked findings (0 for, 2 against)
02Theme
Intervention design
- Guardrailed AI tutors (hints, no direct answers, pedagogical prompting) improve secondary students' performance on assessments taken without AI, compared with business-as-usual instruction.Contradictory · Secondary students · Evidence unclear · 3 linked findings (1 for, 1 against)
- Guardrails prevent the unassisted-performance harm observed when secondary students practice with answer-giving chatbots.Preliminary signal · Secondary students · Evidence stable · 2 linked findings (2 for, 0 against)
- AI feedback on writing produces unassisted writing growth noninferior to human feedback — no worse than a stated margin.Insufficient evidence · University students · Evidence stable · 1 linked finding (0 for, 0 against)
03Theme
Durability and moderators
- Learning gains from AI-assisted practice persist on delayed (one week or longer) retention tests taken without AI.Insufficient evidence · Secondary students · Evidence stable · 1 linked finding (0 for, 0 against)
- Learning gains from AI-assisted practice persist on delayed (one week or longer) retention tests taken without AI.Insufficient evidence · University students · Evidence stable · 1 linked finding (0 for, 1 against)
- Secondary students with low prior knowledge gain less, or are harmed more, by AI assistance during practice than high-prior-knowledge peers.Preliminary signal · Secondary students · Evidence unclear · 3 linked findings (2 for, 0 against)
- University students with low prior knowledge gain less, or are harmed more, by AI assistance during practice than high-prior-knowledge peers.Proxy outcomes only · University students · Evidence unclear · 1 linked finding (0 for, 0 against)