{"ID":2846170,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.02453","arxiv_id":"2511.02453","title":"Accounting for Underspecification in Statistical Claims of Model Superiority","abstract":"Machine learning methods are increasingly applied in medical imaging, yet many reported improvements lack statistical robustness: recent works have highlighted that small but significant performance gains are highly likely to be false positives. However, these analyses do not take \\emph{underspecification} into account -- the fact that models achieving similar validation scores may behave differently on unseen data due to random initialization or training dynamics. Here, we extend a recent statistical framework modeling false outperformance claims to include underspecification as an additional variance component. Our simulations demonstrate that even modest seed variability ($\\sim1\\%$) substantially increases the evidence required to support superiority claims. Our findings underscore the need for explicit modeling of training variance when validating medical imaging systems.","short_abstract":"Machine learning methods are increasingly applied in medical imaging, yet many reported improvements lack statistical robustness: recent works have highlighted that small but significant performance gains are highly likely to be false positives. However, these analyses do not take \\emph{underspecification} into account...","url_abs":"https://arxiv.org/abs/2511.02453","url_pdf":"https://arxiv.org/pdf/2511.02453v1","authors":"[\"Thomas Sanchez\",\"Pedro M. Gordaliza\",\"Meritxell Bach Cuadra\"]","published":"2025-11-04T10:31:34Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"eess.IV\"]","methods":"[]","has_code":false}