SOURCE-LINKED INTELLIGENCE

When More Evidence Hurts: Publication-Bias Drift and Principled Stopping for Biomedical Causal Search

arXiv · AI, language, vision and robotics · article · Sep 21, 2026 · UTC

Automated biomedical evidence synthesis depends on retrieving published studies, but the biomedical literature is systematically skewed toward positive findings. Deeper retrieval can therefore make a system \emph{more} likely to falsely infer benefit when the true effect is null. We formalise this phenomenon as \emph{evidence drift} and prove that, under a standard publication-bias model, the false-positive probability on null-effect queries follows a strictly increasing large-sample envelope in retrieval depth, approaching one. Empirically, on a held-out test set of 140 Cochrane-derived queri

Read original source ↗ Open in workspace

recordType

paper

region

Global