uid stringclasses 10
values | dataset_id stringclasses 1
value | jurisdiction stringclasses 3
values | court_level stringclasses 5
values | year stringclasses 5
values | area_of_law stringclasses 5
values | pillar stringclasses 2
values | series stringclasses 3
values | input_expert_field_and_scope stringclasses 10
values | input_method_summary stringclasses 10
values | input_data_relied_on stringclasses 10
values | input_conclusion_summary stringclasses 10
values | input_confidence_language stringclasses 10
values | input_cross_exam_vulnerability_signals stringclasses 9
values | question stringclasses 3
values | ground_truth_label stringclasses 4
values | ground_truth_rationale stringclasses 10
values | outcome_signal stringclasses 6
values | early_window_months stringclasses 5
values | source_citation stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
ER-TR-001 | legal-expert-report-method-conclusion-coherence-v0.1 | UK | High Court | 2022 | Commercial | legal_actors | 3C | Forensic accounting on loss quantification | Standard discounted cash flow with sensitivity analysis | Audited accounts and bank statements | Loss estimated £2.1m with range | Uses ranges and assumptions clearly | Low vulnerability | Is expert report coherent. Answer coherent or incoherent. | coherent | Method fits field and data supports bounded conclusion. | stable | 0 | Synthetic |
ER-TR-002 | legal-expert-report-method-conclusion-coherence-v0.1 | UK | High Court | 2022 | Commercial | legal_actors | 3C | Forensic accounting | No model explained | Unaudited projections only | Loss stated £5m as certain | Absolute certainty language | High vulnerability. No workings | Is expert report coherent. Answer coherent or incoherent. | incoherent | Conclusion unsupported by method and certainty inflated. | exclusion_risk | 2 | Synthetic |
ER-TR-003 | legal-expert-report-method-conclusion-coherence-v0.1 | US | Federal | 2021 | Products | legal_actors | 3C | Engineering failure analysis | Materials testing and fracture analysis | Lab results and exemplar testing | Defect likely caused failure with probabilities | Qualified conclusions | Addresses alternative causes | Is expert report coherent. Answer coherent or incoherent. | coherent | Method applied and alternatives addressed. | stable | 0 | Synthetic |
ER-TR-004 | legal-expert-report-method-conclusion-coherence-v0.1 | US | Federal | 2021 | Products | legal_actors | 3C | Engineering expert opines on medical causation | No medical method | Engineering notes only | States injury medically caused by product | High certainty | Outside expertise scope | Is expert report coherent. Answer coherent or incoherent. | incoherent | Scope overreach and method mismatch. | daubert_risk | 3 | Synthetic |
ER-TR-005 | legal-expert-report-method-conclusion-coherence-v0.1 | UK | Crown Court | 2020 | Criminal | legal_actors | 3C | Digital forensics on device attribution | Hash verification and timeline reconstruction | Device logs and metadata | User activity consistent with suspect access | Uses probability and limitations | Alternative user scenario assessed | Is expert report coherent. Answer coherent or incoherent. | coherent | Forensic steps support conclusion with limits. | stable | 0 | Synthetic |
ER-TR-006 | legal-expert-report-method-conclusion-coherence-v0.1 | UK | Crown Court | 2020 | Criminal | legal_actors | 3C | Digital forensics | No reproducible steps | Selective screenshots | States definitive attribution to suspect | Absolute language | Ignores alternative access | Is expert report coherent. Answer coherent or incoherent. | incoherent | Non-reproducible method and ignored alternatives. | credibility_risk | 2 | Synthetic |
ER-TR-007 | legal-expert-report-method-conclusion-coherence-v0.1 | EU | Commercial Court | 2019 | Construction | legal_actors | 3C | Delay analysis expert | Critical path method with contemporaneous records | Programmes, site diaries, change orders | Delay apportioned with ranges | Assumptions declared | Alternative scenarios run | Is expert report coherent. Answer coherent or incoherent. | coherent | Method standard and data supports apportionment. | stable | 0 | Synthetic |
ER-TR-008 | legal-expert-report-method-conclusion-coherence-v0.1 | EU | Commercial Court | 2019 | Construction | legal_actors | 3C | Delay expert | Narrative only no CPM | Anecdotal statements | Assigns all delay to defendant | Absolute blame language | No sensitivity analysis | Is expert report coherent. Answer coherent or incoherent. | incoherent | Conclusion not supported by method or data. | exclusion_risk | 4 | Synthetic |
ER-TR-009 | legal-expert-report-method-conclusion-coherence-v0.1 | US | State | Medical Malpractice | legal_actors | 3C | Medical expert on standard of care | Differential analysis with guideline comparison | Records and guidelines | Breach likely contributed to harm | Uses likelihood language | Addresses confounders | Is expert report coherent. Answer coherent or incoherent. | coherent | Method and scope fit with careful probability framing. | stable | 0 | Synthetic | null |
ER-TR-010 | legal-expert-report-method-conclusion-coherence-v0.1 | US | State | Medical Malpractice | legal_actors | 3C | Medical expert | No guideline discussion | Selective records only | States breach definitely caused outcome | Absolute language | Ignores alternative causes | Is expert report coherent. Answer coherent or incoherent. | incoherent | Conclusion overstates causation beyond analysis. | daubert_risk | 3 | Synthetic | null |
What this dataset does
You receive
expert scope method data conclusion confidence language cross-exam vulnerabilities
You decide
coherent or incoherent
Daily use
quick expert QC before filing
daubert attack planning
cross-exam target list
- Downloads last month
- 19