{
  "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) and % change -35.0% at 5 km, and the text states significance at the 5% level, matching every element of the assertion.",
    "A2": "Table 7 prints the 1 km HIV ATT as +0.063 (SE 0.801), positive and insignificant, so the claimed negative sign and the claimed fall in HIV hospitalisations are contradicted.",
    "A3": "The ATT of -0.283 log points corresponds to a -24.7% change (printed in Table 6; 100(exp(-0.283)-1) = -24.65%), not a 28.3% reduction, which wrongly reads log points as percent.",
    "A4": "With ATT -0.406 and SE 0.250 the 95% CI is [-0.896, 0.084] (±1.96·SE), which includes zero (the estimate carries no star); the asserted interval is only ±1 SE.",
    "A5": "The extract reports ATTs only by buffer (Table 6), outcome category (Table 7) and estimator (Table A19) and never by sex, so the men-versus-women comparison can be neither confirmed nor refuted.",
    "A6": "Table 2 prints the ATE as -1.99 weeks (95% CI [-2.24, -1.75]) with a control mean of 17.3 weeks, and 1.99/17.3 = 11.50% matches the 11.5% decrease 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 0.218 in Table A19 (internal inconsistency).",
      "handling": "A1 cites Table 6, whose SE (0.210) matches the assertion; verdict uses Table 6. With SE 0.218, |z| = 1.977: still significant at 5% under 1.96, borderline under 2.0; A1 verdict unchanged because the printed ** appears in both tables."
    },
    {
      "issue": "arxiv-2609.27732v1: Table A19 note says 640 never-treated comparison regions while the Table 6/A19 rows and the Figure 6 note say 608.",
      "handling": "Not decisive for any assertion; recorded only."
    },
    {
      "issue": "arxiv-2609.27732v1: narrative percentages for sub-outcomes are inconsistent: abstract says fatal hospitalisations fall by 35% while Section 6.1 says ~30% at 5 km (100(exp(-0.354)-1) = -29.8%); Section 6.1 says narcotics +47% at 5 km (implied 47.7%); text calls HIV estimates 'close to zero' although the 5 km HIV ATT is -0.340.",
      "handling": "Verdicts rely on the printed table coefficients, not the narrative; none of these is decisive for A1-A6."
    },
    {
      "issue": "arxiv-2606.14621v1: results text reports the weeks-on-UI CIs as positive magnitudes ([1.75 weeks, 2.24 weeks]) for a 'reduction', while Table 2 prints signed [-2.24, -1.75]; also a rendering defect 'p ¡ 0.001' for p < 0.001.",
      "handling": "Treated as equivalent sign conventions; A6 judged against the signed Table 2 values."
    },
    {
      "issue": "Both extracts are partial (sections omitted), and RESEA Tables 2-3 were re-set by the extract editor.",
      "handling": "Verdicts rely only on the extract content as instructed; values in re-set tables are taken as verbatim per the editorial note."
    }
  ],
  "n_input_rows": {
    "assertions.json": 6,
    "README.md": 14,
    "arxiv-2609.27732v1.txt": 134,
    "arxiv-2606.14621v1.txt": 58
  },
  "n_analysed": 6,
  "random_seed": null,
  "supporting_calculations": {
    "A1": {
      "b": -0.431,
      "se": 0.21,
      "stars": 2,
      "pct_printed": -35.0,
      "pct_implied": -35.01,
      "z": -2.052
    },
    "A2": {
      "b": 0.063,
      "se": 0.801,
      "stars": 0
    },
    "A3": {
      "b": -0.283,
      "pct_printed": -24.7,
      "pct_implied": -24.65
    },
    "A4": {
      "b": -0.406,
      "se": 0.25,
      "stars": 0,
      "ci95_1p96": [
        -0.896,
        0.084
      ],
      "ci95_2p00": [
        -0.906,
        0.094
      ],
      "ci_pm_1se": [
        -0.656,
        -0.156
      ]
    },
    "A5": {
      "sex_lines": []
    },
    "A6": {
      "ate": -1.99,
      "ci": [
        -2.24,
        -1.75
      ],
      "control_mean": 17.3,
      "rel_pct": 11.5
    }
  },
  "baseline_verdicts_verbatim_number_match": {
    "A1": "supported",
    "A2": "supported",
    "A3": "not_supported",
    "A4": "not_supported",
    "A5": "supported",
    "A6": "supported"
  },
  "baseline_agreement_with_verdicts": 4
}