Evidence (7278 claims)
Search and filter individual claims pulled from the papers. Looking for a specific finding ("what's the effect on wages?"), you're in the right place. Want to compare whole outcome categories against each other instead? Use the Evidence Explorer.
The board below groups claims two ways: by broad theme (nine paper-level topics) and by outcome category (the 34 claim-level outcomes that the Explorer and Syntheses also use).
Browse by theme
Nine broad, paper-level topics. Click one to filter the claims below.
Adoption
9047 claims
Filter claims →
Productivity
8066 claims
Filter claims →
Governance
7278 claims
Filtered →
Human-AI Collaboration
6912 claims
Filter claims →
Org Design
4439 claims
Filter claims →
Innovation
4359 claims
Filter claims →
Labor Markets
3652 claims
Filter claims →
Skills & Training
3018 claims
Filter claims →
Inequality
2160 claims
Filter claims →
Claims by outcome category
Counts by direction of finding. These are the same 34 outcome categories the Explorer compares and the Syntheses are written for. A linked row has a published synthesis.
| Outcome | Positive | Negative | Mixed | Null | Total |
|---|---|---|---|---|---|
| Other | 795 | 210 | 105 | 955 | 2131 |
| Governance & Regulation | 886 | 414 | 197 | 126 | 1654 |
| Organizational Efficiency | 826 | 204 | 129 | 87 | 1257 |
| Technology Adoption Rate | 681 | 259 | 128 | 110 | 1189 |
| Research Productivity | 464 | 138 | 65 | 349 | 1028 |
| Output Quality | 503 | 196 | 61 | 53 | 813 |
| Decision Quality | 351 | 180 | 84 | 51 | 673 |
| AI Safety & Ethics | 238 | 288 | 71 | 34 | 637 |
| Firm Productivity | 455 | 58 | 92 | 20 | 631 |
| Market Structure | 186 | 172 | 123 | 25 | 511 |
| Task Allocation | 222 | 70 | 76 | 34 | 407 |
| Innovation Output | 238 | 28 | 48 | 18 | 334 |
| Skill Acquisition | 177 | 62 | 62 | 17 | 318 |
| Employment Level | 107 | 57 | 108 | 13 | 287 |
| Fiscal & Macroeconomic | 135 | 72 | 44 | 26 | 284 |
| Firm Revenue | 172 | 50 | 28 | 5 | 256 |
| Consumer Welfare | 121 | 68 | 45 | 12 | 246 |
| Task Completion Time | 183 | 33 | 10 | 13 | 240 |
| Inequality Measures | 45 | 126 | 50 | 6 | 227 |
| Worker Satisfaction | 95 | 74 | 23 | 12 | 204 |
| Error Rate | 77 | 98 | 11 | 4 | 190 |
| Regulatory Compliance | 84 | 73 | 17 | 7 | 181 |
| Automation Exposure | 61 | 61 | 27 | 14 | 166 |
| Training Effectiveness | 98 | 21 | 14 | 19 | 154 |
| Wages & Compensation | 78 | 37 | 25 | 6 | 146 |
| Developer Productivity | 105 | 18 | 14 | 6 | 144 |
| Team Performance | 87 | 17 | 28 | 10 | 143 |
| Job Displacement | 12 | 83 | 23 | 1 | 119 |
| Hiring & Recruitment | 53 | 8 | 8 | 3 | 72 |
| Social Protection | 39 | 17 | 8 | 2 | 66 |
| Creative Output | 32 | 20 | 8 | 3 | 64 |
| Skill Obsolescence | 5 | 50 | 6 | 1 | 62 |
| Labor Share of Income | 17 | 20 | 17 | — | 54 |
| Worker Turnover | 15 | 15 | — | 3 | 33 |
| Industry | — | — | — | 1 | 1 |
Governance
Remove filter
Management principles emphasised are transparency, traceability of outcomes, IT integration for documentation, and continuous monitoring/evaluation.
Explicit management principles in paper (prescriptive).
Research and audit should emphasise validity, reliability, and compliance using mixed methods (qualitative interviews/focus groups; quantitative surveys/statistics) and systematic curriculum audits.
Recommended research & audit approach in paper (methodological guidance).
Tools recommended include logigrams (visual decision/compliance flows) and algorigram (algorithmic step-flows for planning, assessment, audit).
Tool definitions and recommendations in paper (descriptive).
Core components of the framework are inputs (learner needs, industry requirements, regulatory standards), processes (curriculum mapping, competency alignment, career assessment), and outputs (structured lesson plans, compliance-ready frameworks, career-path documentation).
Framework component list provided in paper (descriptive).
Scope of the program includes curriculum design, organisational management, career-alignment, and audit/compliance processes.
Explicit scope statement in paper (descriptive).
The framework foregrounds logical modelling (logigrams, algorigrams) and mixed-methods data analysis to support design, auditability, and alignment with industry and regulatory standards.
Paper's methodological design and tool recommendations (conceptual). No empirical implementation data reported.
The program offers a comprehensive curriculum-engineering framework linking organizational orientation, management systems, lesson planning, and career assessment into traceable, compliance-ready curriculum products.
Paper's program description and framework specification (conceptual); no empirical evaluation or sample size reported.
The paper calls for subsequent quantitative validation (using task-based, matched employer-employee, and provider-level panel data) to estimate causal impacts on productivity, health outcomes, wages, and employment composition across the three interaction levels.
Stated research agenda and measurement recommendations in the paper's discussion section.
The study is qualitative and small-sample (four case) and therefore interpretive and illustrative rather than statistically generalizable.
Explicit methodological statement in the paper: design = qualitative multiple case study, sample = four AI healthcare applications.
The study identifies a three-level taxonomy of human–AI interaction in healthcare: AI-assisted, AI-augmented, and AI-automated.
Conceptual taxonomy derived from multiple qualitative case studies (n=4) using cross-case comparison and Bolton et al. (2018)'s three-dimensional service-innovation framework.
Non-probability sampling and self-reported measures limit claims about prevalence and causality; cross-sectional design cannot capture dynamics of skill acquisition over time.
Study limitations explicitly reported by authors: non-probability sampling, self-reported measures, and cross-sectional design.
There are few large-scale randomized controlled trials (RCTs) showing direct patient outcome improvements from GenAI CDS; high-quality real-world and longitudinal studies are limited but essential.
Evidence-maturity statement in the paper summarizing the literature; the paper explicitly notes scarcity of large RCTs and longitudinal evaluations.
Randomized or quasi-experimental evaluations of digital-ID rollouts, subsidy programs for fintech adoption, or sandboxed regulatory innovations can identify causal impacts on inclusion and growth.
Methodological recommendation proposing experimental and quasi-experimental designs to obtain causal inference; no implementation results reported in the paper summary.
AI economists should prioritize measuring how AI-driven services affect access, default rates, transaction costs, and market structure, disaggregated across income groups and regions.
Methodological recommendation in the 'Implications for AI Economics' section; suggested measurement priorities rather than an empirical finding.
There is a need for economic analysis of data governance regimes, model transparency requirements, algorithmic auditability, and incentives for responsible AI adoption in finance.
Methodological and policy recommendation based on identified gaps in the literature and regulatory practice; this is a stated research/policy need in the paper rather than an empirical claim requiring sample evidence.
Typical evaluation metrics reported are accuracy, precision, recall, F1-score, AUC, detection rate, false positive rate, latency, and computational cost.
Survey of evaluation practices in reviewed papers listing the metrics authors commonly report.
Emerging approaches in the literature include federated learning, online/streaming learning, and transfer learning for cross-device generalization.
Trend analysis across recent papers indicating adoption of federated and continual learning paradigms and transfer-learning techniques.
Unsupervised and semi-supervised methods (clustering, one-class classifiers, autoencoder-based anomaly detectors) are commonly employed to handle unlabeled/anomalous IoT traffic.
Synthesis of studies using anomaly-detection paradigms and unsupervised techniques reported in the reviewed papers.
Deep learning approaches used include CNNs, RNNs/LSTMs for sequence/traffic analysis, and autoencoders for anomaly detection.
Surveyed literature and taxonomy noting multiple studies that apply convolutional and recurrent architectures and autoencoders to network/traffic data.
Common ML approaches reported for IoT IDS include supervised models (random forest, SVM, gradient boosting, neural networks).
Taxonomy and literature synthesis showing frequent use of classical supervised classifiers in surveyed papers and experiments.
This work is a conceptual framework and design proposal synthesizing methods from recommender systems and HRI rather than a report of novel empirical experiments.
Explicit statement in the Data & Methods section of the paper.
The review followed PRISMA guidelines and included 30 scholarly articles retrieved from Scopus, published between 2020 and 2025, selected using pre-specified inclusion criteria.
Methods section of the paper reporting the SLR protocol, database, time window, and number of included studies.
The study is primarily diagnostic and prescriptive rather than empirical: no explicit empirical dataset, causal identification strategy, or statistical estimation is reported.
Methods section of the paper explicitly characterizes the work as conceptual, systems-oriented, and not reporting empirical evaluation data.
The study's empirical identification relies on longitudinal variation with city fixed effects and time effects, plus non-linear/threshold identification via polynomial (DE^2) terms and threshold-regression using green-technology-innovation as the threshold variable.
Description of empirical strategy in the paper: panel fixed-effects models (controlling for time-invariant city heterogeneity and common time shocks), mediating-effect models for channel tests, and threshold-regression models for regime-dependent effects, applied to the 278-city 2011–2022 panel.
Research recommendation: invest in longer-run, rigorous impact evaluations (RCTs, panel studies) and system-level assessments to capture spillovers and sustainability outcomes.
Authors' stated research agenda based on identified methodological gaps (limited long-term and system-level evidence) in the review.
There is variation in study design and quality in the evidence base (RCTs, quasi-experimental studies, observational case studies, pilots).
Methodological caveats noted by the authors summarizing the diversity of designs reported across reviewed studies.
The review used a structured literature review with thematic synthesis and a comparative effect-size analysis to quantify ranges for yield, cost, and efficiency outcomes.
Authors' description of review approach and analytical methods in the Data & Methods section.
The evidence base reviewed comprises more than 60 peer-reviewed articles and institutional reports from 2020–2025, primarily focusing on Sub-Saharan Africa.
Statement in the paper's Data & Methods section describing the scope and composition of the review sample.
Effect sizes and impacts vary substantially across contexts—by crop, farm size, and institutional setting.
Comparative synthesis across studies showing heterogeneity in reported outcomes and authors' methodological caveats highlighting context dependence.
Technologies assessed in the review include predictive analytics, digital advisory systems, smart irrigation, pest/disease detection, and precision fertilization.
Descriptive synthesis of the types of AI and digital technologies evaluated across the >60 reviewed articles and reports (2020–2025).
These quantitative performance figures come from case‑level, high‑performer pilots and should not be treated as typical industry benchmarks.
Authors' caveat based on the composition of evidence in the review (skew towards pilots and selected advanced implementations; limited longitudinal/multi‑project empirical studies).
Inter‑rater reliability for the study selection/encoding was Cohen’s κ = 0.83 (substantial agreement).
Reported inter‑rater reliability statistic from the review's quality control step (Cohen's kappa = 0.83).
The review screened 463 Scopus records (2018–2026) and selected 160 peer‑reviewed studies using a PRISMA‑guided process.
Systematic literature review described in paper: Scopus search (2018–2026), PRISMA screening and eligibility filtering; initial n=463, final n=160.
Kebumen UNESCO Global Geopark is used as a practical context to ground the framework; its ecological/cultural assets and emergent digital presence make it a suitable case for studying emerging destinations balancing innovation with authenticity.
Paper provides Kebumen Geopark as the illustrative case study/context for the conceptual framework; no systematic case-study data reported.
Operationalization suggestions: social proof via ratings, reviews, UGC volume and valence; behavioral proxies include bookings and inquiries as outcomes.
Paper explicitly lists social-proof indicators and behavioral proxies as part of recommended empirical approaches (digital-trace and platform data).
Operationalization suggestions: sustainability communication via message clarity, perceived authenticity, and specificity of eco-actions.
Operationalization guidance in the paper for measuring sustainability messaging in experiments/surveys.
Operationalization suggestions: AI personalization via perceived relevance, transparency, and perceived fairness of recommendations.
Operationalization guidance in the paper; proposed as latent construct indicators for future SEM or experiments.
Operationalization suggestions: digital experience quality via usability, information richness, responsiveness, multi-channel integration.
Operationalization guidance provided in the paper's methods suggestions; intended for future empirical measurement.
Recommended empirical follow-ups include Structural Equation Modeling (SEM), experimental tests (lab/field/online), quasi-experimental causal-inference methods (DiD, IVs, RD), comparative/regional designs, and analysis of digital-trace/platform data (clickstreams, recommendation logs, bookings, UGC).
Methodological recommendations explicitly listed in the Data & Methods and Research Agenda sections of the paper; no primary empirical work conducted.
The framework produces ten testable propositions mapping hypothesized direct and mediated links among constructs and specifying contingencies for future empirical testing.
Explicit statement in the paper that the framework yields ten testable propositions; no empirical validation reported.
Experimental structure determination (X‑ray, NMR, cryo‑EM) remains the gold standard but is slow, costly, and low‑throughput.
Paper explicitly states experimental methods are 'gold standard' and characterizes them as slow, costly, low‑throughput; the PDB is cited as the source of structural ground truth.
The authors did not perform primary empirical validation or simulation of TVR‑Sec across real VR deployments.
Methods and limitations section explicitly state no original empirical experiments or simulations were conducted; analysis is conceptual and qualitative.
The paper's scope comprised a comparative literature review and conceptual integration of 31 peer‑reviewed studies published between 2023 and 2025.
Authors' methods description specifying sample size and publication window: 31 peer‑reviewed studies (2023–2025).
This study is descriptive and comparative rather than quantitative; it relies on available policy documents and secondary literature rather than original field interviews or measured outcomes.
Explicit methodological statement in the paper listing qualitative document analysis, comparative literature review, and policy commentary; limitation acknowledged by authors.
A research agenda for AI economics should include: formalizing consent as a transaction/contracting problem; empirical RCTs and natural experiments measuring effects of consent designs; mechanism design for privacy-preserving data sharing; and policy evaluation of consent regulations.
Explicitly listed research directions in the workshop outputs and position papers; these are proposed next steps rather than empirical findings.
Follow-up empirical methods should include qualitative interviews, focus groups, usability studies, field experiments (A/B tests), and policy/legal-technical assessments.
Recommended research methods enumerated in the workshop outputs and position papers; these are proposed future methods rather than findings from conducted studies.
The Futures Design Toolkit (scenario planning, persona generation, speculative design) was used as a primary method in the workshop.
Methodological description in the workshop summary listing the Futures Design Toolkit and associated activities; procedural claim rather than empirical.
Empirical generalization across all climate-AI systems is constrained by heterogeneous data availability and proprietary models, limiting the ability to produce universal quantitative claims.
Stated methodological limitation in the paper, noting heterogeneous data and the proprietary nature of some models restrict broad generalization.
The paper does not provide granular quantitative estimates of the economic cost of infrastructural asymmetries in climate-AI.
Explicit limitation stated by the authors in the Methods/Limitations section.
There is a need for empirical research quantifying earnings dispersion, labor substitution effects, and the welfare impacts of GenAI-driven content economies over time.
Explicit research recommendation made in the paper based on gaps identified during analysis of the 377 videos (study is qualitative and does not measure these outcomes).