1{2 "task_id": "F4-conf-02",3 "verdicts": {4 "A1": "supported",5 "A2": "not_supported",6 "A3": "not_supported",7 "A4": "not_supported",8 "A5": "not_assessable",9 "A6": "supported"10 },11 "locators": {12 "A1": "arxiv-2609.27732v1 | Table 6 | row \"ATT\" | column \"5 km\"",13 "A2": "arxiv-2609.27732v1 | Table 7 | row \"HIV\" | column \"1 km\"",14 "A3": "arxiv-2609.27732v1 | Table 6 | row \"% change\" | column \"3 km\"",15 "A4": "arxiv-2609.27732v1 | Table 7 | row \"Hepatitis\" | column \"5 km\"",16 "A5": "not reported in arxiv-2609.27732v1",17 "A6": "arxiv-2606.14621v1 | Table 2 | row \"Weeks on UI\" | column \"Average treatment effect estimate\""18 },19 "reasons": {20 "A1": "Table 6 prints ATT -0.431** with SE 0.210 (** = p<0.05) and % change -35.0%, and 100(exp(-0.431)-1) = -35.01%, so every element matches.",21 "A2": "Table 7 prints the 1 km HIV ATT as +0.063 (SE 0.801), positive and insignificant, so the claimed negative sign and implied fall are contradicted.",22 "A3": "Table 6 converts the 3 km ATT -0.283 to -24.7% (100(exp(-0.283)-1) = -24.65%), not 28.3%; the assertion reads log points as percent (and the estimate is insignificant, SE 0.279).",23 "A4": "With ATT -0.406 and SE (0.250) the 95% CI is [-0.896, 0.084] (±2SE: [-0.906, 0.094]), which includes zero; the asserted [-0.66, -0.16] is only ±1 SE ([-0.656, -0.156]) and the cell carries no significance star.",24 "A5": "The extract reports ATTs only by buffer (Table 6), outcome category (Table 7) and estimator (Table A19), never by sex, so the men-vs-women comparison can be neither confirmed nor refuted.",25 "A6": "Table 2 prints the ATE as -1.99 weeks (95% CI [-2.24, -1.75]) against a control mean of 17.3 weeks, and 1.99/17.3 = 11.5%, matching every element."26 },27 "n_supported": 2,28 "n_not_supported": 3,29 "n_not_assessable": 1,30 "feasibility": "complete",31 "feasibility_reason": null,32 "data_issues": [33 {34 "issue": "arxiv-2609.27732v1: Table A19 CS-DR column prints SE (0.218) for the ATT -0.431 that Table 6 prints with SE (0.210) (same estimator, buffer and sample).",35 "handling": "A1 decided on Table 6, which the assertion quotes; under SE 0.218 the estimate remains significant at 5% (|t|=1.98 > 1.96) and carries ** in both tables, so the verdict is unaffected."36 },37 {38 "issue": "arxiv-2609.27732v1: Table A19 note states 640 never-treated comparison regions whereas Tables 6, 7 and A19 body state 608.",39 "handling": "Recorded only; no assertion depends on the control-region count."40 },41 {42 "issue": "arxiv-2609.27732v1: abstract says fatal hospitalisations fall by 35%, but Table 7 5 km fatal ATT -0.354 converts to -29.8% (Section 6.1 says ~30%); abstract appears to read log points as percent.",43 "handling": "Recorded only; verdicts rely on table cells, not the abstract; no assertion concerns fatal outcomes."44 },45 {46 "issue": "arxiv-2609.27732v1: results text calls HIV estimates 'close to zero' although the 5 km HIV ATT is -0.340 (SE 0.294).",47 "handling": "Recorded only; A2 decided on the printed 1 km cell (0.063)."48 },49 {50 "issue": "arxiv-2606.14621v1: text reports CIs for reductions as positive magnitudes ([1.75 weeks, 2.24 weeks]) while Table 2 prints [-2.24, -1.75]; also a rendering defect 'p ¡ 0.001'.",51 "handling": "Treated as sign convention of the prose; A6 decided on Table 2, whose signed interval matches the assertion."52 }53 ],54 "n_input_rows": {55 "assertions.json": 6,56 "README.md": 14,57 "arxiv-2609.27732v1.txt": 134,58 "arxiv-2606.14621v1.txt": 5859 },60 "n_analysed": 6,61 "random_seed": null,62 "baseline_naive_number_match_verdicts": {63 "A1": "supported",64 "A2": "supported",65 "A3": "not_supported",66 "A4": "not_supported",67 "A5": "supported",68 "A6": "supported"69 },70 "baseline_agreement_with_final": 4,71 "derived": {72 "A1_pct_from_logpoints": -35.01,73 "A1_t_stat_table6": -2.052,74 "A3_pct_from_logpoints": -24.65,75 "A4_ci95_1.96": [76 -0.896,77 0.08478 ],79 "A4_ci95_2.0": [80 -0.906,81 0.09482 ],83 "A4_ci_1se": [84 -0.656,85 -0.15686 ],87 "A6_relative_reduction_pct": 11.588 }89}