study,estimate_as_printed,uncertainty_as_printed,sample_as_printed,definition,locator,harmonised_estimate,harmonised_se,harmonisable,poolable,transformation,reason_if_not
S1,"ExpCon coefficient -0.24. (Model 2, logged time on task); text: 'effect of AI = 21%' faster, 'about 21%' in abstract","SE (0.14); 95% CI = [-0.51, 0.03]; p=0.086 (NS); marker '.' p<.1","n = 96 (Table III); N=47 AI, N=46 no-AI (Sec VI-B)","OLS coefficient of the AI-exposure indicator (ExpCon) on log time on task (LogToT), Model 2 with developer-level covariates; one complex enterprise coding task, randomised AI vs no-AI","arxiv-2410.12944v3, Table III, row ExpCon (+ SE row below), column Model (2) 'Dev'; Sec VI-C sentence 'estimate of experimental condition on best-fit Model 2'",-0.24,0.14,True,True,theta = b = -0.24 (log-time coefficient); SE(theta) = SE(b) = 0.14 (printed). Cross-check from CI: (0.03-(-0.51))/(2*1.96) = 0.1378,
S2,+19% (allowing AI increases completion time by 19%; 'implement issues in 19% more time on average'),"none printed (95% CI shown only in Figure 1 / Figure 15, not reproduced in the extract)","16 developers, 246 tasks (136 AI-allowed, 110 AI-disallowed issues)","S = E[time | AI allowed]/E[time | AI disallowed] - 1 = exp(beta) - 1, beta from OLS of log implementation time on AI indicator and log forecast time; issue-level randomisation","arxiv-2507.09089v2, Abstract and Section 3 opening paragraph ('19% more time'); estimator Sec 2.3 and App. D.1-D.3",0.173953,,True,False,"theta = ln(1 + c) = ln(1 + 0.19) = 0.173953 (equivalently theta = beta, since S = exp(beta) - 1)","Not poolable: no SE, CI or p-value for the 19% estimate is printed in the extract; the confidence interval appears only in Figure 1 (not reproduced), so SE(theta) cannot be obtained."
S3,"All Content: Overall 53.84, Treatment 44.29, Control 63.10 (minutes), Improvement -29.79%***",significance stars only (*** = 0<p<=0.01 per the paper's definition); no SE or CI printed,181 subjects (treatment/control split ~half each; per-arm N not printed in extract),"difference in task duration (minutes) between Copilot and control groups, holding accuracy constant (control duration predicted at treatment accuracy, App. B), all scenarios and tasks","arxiv-2411.01067v2, Section 4.2, Table 3, block 'Aggregate Results', row 'All Content', columns Treatment / Control / Improvement",-0.353679,,True,False,theta = ln(1 + c) = ln(1 - 0.2979) = -0.353679 (ratio of printed times: ln(44.29/63.10) = -0.353962),"Not poolable: only significance stars (p<=0.01) are printed; no SE, CI or exact p-value, so SE(theta) cannot be obtained."
S4,"Full Corpus: Aware Time per Email (s) 50.5, Control Time per Email (s) 56.9, % Time Savings 11.1",p-value 0.471,"Aware N=56, Control N=52 (167 analysts across three arms after filtering)","percentage time saving (Control vs Aware arm) in email-processing time over the full 25-email queue, without the resolve-benign protocol; Aware arm sees the agent's verdicts/rationales and an agent-ordered queue","arxiv-2511.13860v1, Section 3.3, Table 3, column 'Full Corpus', rows Aware Time / Control Time / % Time Savings / p-value",-0.117658,0.163221,True,True,time saving 11.1% => c = -0.111; theta = ln(1 - 0.111) = -0.117658; approximate SE = |theta| / z_(1-0.471/2) = 0.117658/0.720853 = 0.163221 (p assumed two-sided). APPROXIMATE SE.,
S5,"Treat x Post -0.278 (Output, column 3, panel A)",SE (0.295); no stars,"N 18332 (user-day observations, users with affiliation)","difference-in-differences coefficient of Italy x post-ban on the aggregate Output measure (count-based GitHub activity: pushes, PRs, comments, creates, issues), pre 27/28.03 vs post 03/04.04 2023, user FE","arxiv-2304.09339v1, Section 4, Table 1, panel A, row 'Treat x Post' (+ SE row), column (3) Output, N row",,,False,False,none (not harmonisable),"Not harmonisable: the effect is on output counts (GitHub events), not on task-completion time for the same tasks, and the contrast is a non-randomised difference-in-differences around the withdrawal of ChatGPT access (ban), not a randomised AI vs no-AI comparison."