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** (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.",21 "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.",22 "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).",23 "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.",24 "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.",25 "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."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: 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').",35 "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)."36 },37 {38 "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.",39 "handling": "Recorded only; no assertion depends on the control-region count."40 },41 {42 "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%'.",43 "handling": "Recorded only; no assertion concerns fatal hospitalisations. Tables preferred over prose where needed."44 },45 {46 "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.",47 "handling": "Recorded only; not decisive for any assertion."48 },49 {50 "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.",51 "handling": "Recorded only; A2 is decided from the printed 1 km coefficient's sign."52 },53 {54 "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'.",55 "handling": "Treated as equivalent descriptions of a reduction; A6 judged against the signed Table 2 values."56 },57 {58 "issue": "arxiv-2606.14621v1: Tables 2 and 3 were re-set by the extract editor (one row per outcome) per the extract header.",59 "handling": "Values used as printed; Table 3 'Difference in means' column duplicates Table 2 ATE column, consistent (-1.99; [-2.24, -1.75])."60 }61 ],62 "n_input_rows": {63 "assertions.json": 6,64 "README.md": 14,65 "sources/arxiv-2609.27732v1.txt": 134,66 "sources/arxiv-2606.14621v1.txt": 5867 },68 "n_analysed": 6,69 "random_seed": null,70 "checks": {71 "A1": {72 "b": -0.431,73 "se": 0.21,74 "stars": 2,75 "pct_printed": "-35.0%",76 "pct_calc": -35.01,77 "z": 2.052,78 "z_A19_se": 1.97779 },80 "A2": {81 "b": 0.063,82 "se": 0.80183 },84 "A3": {85 "b": -0.283,86 "pct_printed": "-24.7%",87 "pct_calc": -24.6588 },89 "A4": {90 "b": -0.406,91 "se": 0.25,92 "stars": 0,93 "ci_196": [94 -0.896,95 0.08496 ],97 "ci_200": [98 -0.906,99 0.094100 ],101 "ci_1se": [102 -0.656,103 -0.156104 ]105 },106 "A5": {107 "sex_mentions": 0108 },109 "A6": {110 "ate": -1.99,111 "ci": [112 -2.24,113 -1.75114 ],115 "control_mean": 17.3,116 "pct_calc": -11.5117 }118 }119}