{
  "task_id": "F4-conf-02",
  "verdicts": {
    "A1": "supported",
    "A2": "not_supported",
    "A3": "not_supported",
    "A4": "not_supported",
    "A5": "not_assessable",
    "A6": "supported"
  },
  "locators": {
    "A1": "arxiv-2609.27732v1 | Table 6 | row \"ATT\" | column \"5 km\"",
    "A2": "arxiv-2609.27732v1 | Table 7 | row \"HIV\" | column \"1 km\"",
    "A3": "arxiv-2609.27732v1 | Table 6 | row \"% change\" | column \"3 km\"",
    "A4": "arxiv-2609.27732v1 | Table 7 | row \"Hepatitis\" | column \"5 km\"",
    "A5": "not reported in arxiv-2609.27732v1",
    "A6": "arxiv-2606.14621v1 | Table 2 | row \"Weeks on UI\" | column \"Average treatment effect estimate\""
  },
  "reasons": {
    "A1": "Table 6 prints ATT -0.431** (SE 0.210) at 5 km with p<0.05 and '% change' -35.0%, and 100*(exp(-0.431)-1) = -35.01%, so sign, magnitude, SE, significance and percentage all match.",
    "A2": "The 1 km HIV ATT in Table 7 is printed as +0.063 (SE 0.801), not -0.063, so the asserted negative sign and the claimed fall in HIV hospitalisations are contradicted.",
    "A3": "A log-point ATT of -0.283 implies 100*(exp(-0.283)-1) = -24.65%, matching the printed '% change' of -24.7%, not 28.3% (the assertion equates log points with percent).",
    "A4": "With ATT -0.406 and SE 0.250 the 95% CI is [-0.896, 0.084] (±1.96·SE) and includes zero (estimate is unstarred); the asserted [-0.66, -0.16] is only ±1 SE.",
    "A5": "The extract reports no estimates by sex (no men/women/gender subgroup appears in the abstract, Section 6, Tables 6-7 or Table A19), so the comparison cannot be confirmed or refuted.",
    "A6": "Table 2 prints the ATE on weeks on UI as -1.99 weeks (p<0.001; 95% CI [-2.24, -1.75]) with control mean 17.3 weeks, and 1.99/17.3 = 11.5%, matching the 11.5% stated in the results text."
  },
  "n_supported": 2,
  "n_not_supported": 3,
  "n_not_assessable": 1,
  "feasibility": "complete",
  "feasibility_reason": null,
  "data_issues": [
    {
      "issue": "arxiv-2609.27732v1: the same CS-DR 5 km ATT (-0.431**) has SE 0.210 in Table 6 but SE 0.218 in Appendix Table A19 (column 'CS-DR').",
      "handling": "A1 judged against Table 6, which the assertion paraphrases; the verdict is unaffected because |t| = 2.052 (SE 0.210) or 1.977 (SE 0.218) both exceed 1.96 and both tables print ** (p<0.05)."
    },
    {
      "issue": "arxiv-2609.27732v1: Table A19 note says 640 never-treated comparison regions while its own body and Tables 6-7 report 608 control regions.",
      "handling": "Recorded only; no assertion depends on the control-region count."
    },
    {
      "issue": "arxiv-2609.27732v1: abstract says fatal outcomes 'fall by 35%', but Table 7 5 km fatal ATT -0.354 implies -29.8%, which Section 6.1 reports as 'approximately 30%'.",
      "handling": "Recorded only; no assertion concerns fatal hospitalisations. Tables preferred over prose where needed."
    },
    {
      "issue": "arxiv-2609.27732v1: Table A19 '% change' row prints -30.3% (Synthetic DiD, -0.362) and -13.9% (Stacked DiD, -0.149) whereas 100*(exp(b)-1) gives -30.37% and -13.84%; differences of 0.1 pp, likely rounding from unrounded coefficients.",
      "handling": "Recorded only; not decisive for any assertion."
    },
    {
      "issue": "arxiv-2609.27732v1: the text describes HIV estimates as 'close to zero' although the 5 km HIV ATT is -0.340 (SE 0.294); they are insignificant but not small in magnitude.",
      "handling": "Recorded only; A2 is decided from the printed 1 km coefficient's sign."
    },
    {
      "issue": "arxiv-2606.14621v1: results text reports weeks-on-UI CIs as magnitudes ([1.75 weeks, 2.24 weeks], [1.55, 1.99]) while Table 2 prints signed intervals [-2.24, -1.75]; and 'p ¡ 0.001' is an HTML rendering defect for 'p < 0.001'.",
      "handling": "Treated as equivalent descriptions of a reduction; A6 judged against the signed Table 2 values."
    },
    {
      "issue": "arxiv-2606.14621v1: Tables 2 and 3 were re-set by the extract editor (one row per outcome) per the extract header.",
      "handling": "Values used as printed; Table 3 'Difference in means' column duplicates Table 2 ATE column, consistent (-1.99; [-2.24, -1.75])."
    }
  ],
  "n_input_rows": {
    "assertions.json": 6,
    "README.md": 14,
    "sources/arxiv-2609.27732v1.txt": 134,
    "sources/arxiv-2606.14621v1.txt": 58
  },
  "n_analysed": 6,
  "random_seed": null,
  "checks": {
    "A1": {
      "b": -0.431,
      "se": 0.21,
      "stars": 2,
      "pct_printed": "-35.0%",
      "pct_calc": -35.01,
      "z": 2.052,
      "z_A19_se": 1.977
    },
    "A2": {
      "b": 0.063,
      "se": 0.801
    },
    "A3": {
      "b": -0.283,
      "pct_printed": "-24.7%",
      "pct_calc": -24.65
    },
    "A4": {
      "b": -0.406,
      "se": 0.25,
      "stars": 0,
      "ci_196": [
        -0.896,
        0.084
      ],
      "ci_200": [
        -0.906,
        0.094
      ],
      "ci_1se": [
        -0.656,
        -0.156
      ]
    },
    "A5": {
      "sex_mentions": 0
    },
    "A6": {
      "ate": -1.99,
      "ci": [
        -2.24,
        -1.75
      ],
      "control_mean": 17.3,
      "pct_calc": -11.5
    }
  }
}