{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-10-17-019",
    "source_story_id": "tmp-brief-rating-scale-sensitivity",
    "edition_id": "mp-2026-10-17-morning-0099",
    "edition_url": "https://themachinepress.com/edition/2026-10-17",
    "position": 21,
    "story_type": "ticker",
    "section": "benchmarks-evals",
    "editorial_classification": "editorial",
    "headline": "Change the Rating Scale, Change One in Five Labels",
    "slug": "change-the-rating-scale-change-one-in-five-labels",
    "dek": "Across five models and three social-science tasks, one scale-design change moved normalized scores by about 0.10 on average.",
    "summary": "Across five models and three social-science tasks, one scale-design change moved normalized scores by about 0.10 on average.",
    "body_text": "Roughly 20 percent of texts changed category across alternative formats, ranges or signs, with the largest sensitivity on toxicity. Rubric prompting did not remove the effect. The finding argues for robustness checks; it does not invalidate every model-based score.",
    "why_it_matters": "Across five models and three social-science tasks, one scale-design change moved normalized scores by about 0.10 on average.",
    "limitations": [
      "Rubric prompting did not remove the effect.",
      "The finding argues for robustness checks; it does not invalidate every model-based score."
    ],
    "importance": 6,
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-17-019/change-the-rating-scale-change-one-in-five-labels",
    "json_url": "https://themachinepress.com/story/mp-2026-10-17-019.json",
    "first_published_at": "2026-10-17T09:00:00.000-04:00",
    "modified_at": "2026-10-17T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-10-17-019-001",
        "text": "Across five models and three social-science tasks, one scale-design change moved normalized scores by about 0.10 on average.",
        "source_ids": [
          "source-2026-10-17-021"
        ],
        "qualification": "Rubric prompting did not remove the effect."
      }
    ],
    "source_ids": [
      "source-2026-10-17-021"
    ],
    "tags": [
      "measurement",
      "text scoring",
      "robustness"
    ],
    "image_url": null,
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-10-17-021",
      "title": "arXiv preprint 2610.22187, original v1",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2610.22187v1",
      "canonical_url": "https://arxiv.org/abs/2610.22187v1",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-10-15T12:42:15.000-04:00",
      "accessed_at": "2026-10-17T08:26:35.645-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-10-17-019-001"
      ]
    }
  ],
  "corrections": [],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "Change the Rating Scale, Change One in Five Labels",
    "publisher": "The Machine Press",
    "published_at": "2026-10-17T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-17-019/change-the-rating-scale-change-one-in-five-labels"
  }
}