# Contract — F4-conf-01

Question. Extract, exactly as printed, specified treatment-effect point estimates, SEs, CI bounds and counts from two arXiv extracts (Armendariz 2026 DCR staggered DiD, Tables 6/7; Li et al. 2026 RESEA RCT, Table 2), attach locators, derive (a) a 95% CI for the Table 6 5 km ATT from its SE and (b) an SE for the Table 2 wage ATE from its 95% CI, stating the convention, and flag the quantity not reported (exact p-value of HIV 5 km ATT). Estimands. The printed numbers themselves (nothing re-estimated). Derived quantities are meaningful under approximate normality of the estimator (symmetric Wald interval, critical value 1.96). Outputs. out/results.json (keys listed in TASK.md: 11 DCR printed values, dcr_hiv_att_5km_p_value (null if not printed), dcr_hiv_att_5km_significant_10pct, derived dcr_att_5km_ci_low/high, ri_wage_ate + CI, ri_wage_ate_se_reported, derived ri_wage_ate_se_derived, ri_control_mean_wage, ri_reemployment_ate_pp, ri_weeks_ui_ate + CI, locators, derivation_notes, task_id, feasibility, feasibility_reason, data_issues, n_input_rows, n_analysed, random_seed=null); out/report.md (6 sections); out/analysis.py; out/claims.json. Information rules. Only the two extracts are evidence; values as printed (sign, decimals) from the table named in the key; derived keys use printed values of the named table with stated convention; locator format <source id> | Table <n> | row "<..>" | column "<..>"; unreported → null + "not reported in ". Invalidity. Using memory/other versions; using a value from a different table (e.g. Table A19 SE 0.218 instead of Table 6 SE 0.210; Table 3 instead of Table 2; text instead of table); rounding; fabricating an unprinted p-value.