Skip to content

R0050/2026-03-31-02

Research R0050 — Journalism Disciplines
Mode Query
Run date 2026-03-31
Queries 3
Prompt ai-research-methodology 1.1.0
Model Claude Opus 4.6 (1M context)

Three queries investigating whether journalism and adjacent disciplines have formal truth-seeking methodologies with structured evidence evaluation, and how the Wardle-Derakhshan information disorder taxonomy has been adopted in practice.

Queries

Q001 — Journalistic fact-checking frameworks — Very likely (80-95%)

Query: Does any journalistic fact-checking framework include a hierarchical evidence quality scale, calibrated uncertainty language, structured bias assessment domains, or formal source reliability tiering?

Verdict: No framework includes all four features. NewsGuard comes closest with outlet-level source reliability tiering.

Hypothesis Status Probability
H1: All four features present Eliminated
H2: Partial features across frameworks Supported Very likely (80-95%)
H3: No structural features Eliminated

Confidence: High · Sources: 8 · Searches: 3

Full analysis

Q002 — Truth-seeking disciplines survey — High confidence

Query: Beyond intelligence analysis and science, which other disciplines have formal truth-seeking methodologies with structured evidence evaluation?

Verdict: All eight named disciplines are formal. Five contribute novel concepts: law, auditing, FMEA, historical source criticism, SIFT.

Hypothesis Status Probability
H1: All formal + most novel Supported
H2: Formal but mostly redundant Inconclusive
H3: Not formal or not novel Eliminated

Confidence: High · Sources: 9 · Searches: 5

Full analysis

Q003 — Wardle-Derakhshan taxonomy — Very likely (80-95%)

Query: Has the Wardle-Derakhshan Information Disorder Taxonomy been integrated into any formal methodology as a structured classification tool?

Verdict: Remains conceptual. Widely adopted as vocabulary (EU, UNESCO, 37+ newsrooms) but not operationalized into classification procedures.

Hypothesis Status Probability
H1: Fully operationalized Eliminated
H2: Vocabulary adoption, not procedural Supported Very likely (80-95%)
H3: No meaningful adoption Eliminated

Confidence: High · Sources: 5 · Searches: 3

Full analysis


Collection Analysis

Cross-Cutting Patterns

Pattern Queries Affected Significance
Journalism prefers principle-based over structural methodology Q001, Q003 Journalism's professional culture values editorial judgment over formalized scoring — this is a deliberate philosophical choice, not merely an oversight
Vocabulary adoption outpaces procedural adoption Q001, Q003 Both journalism frameworks (Q001) and the information disorder taxonomy (Q003) show high vocabulary adoption without procedural operationalization
Consequence-driven formalization Q001, Q002 Disciplines with legal/financial consequences (law, auditing, engineering safety) have the most formalized evidence evaluation; journalism, with lower consequence for methodological failure, has less
Multiple disciplines converge on similar challenges Q002 Evidence evaluation across disciplines faces the same fundamental problems: source reliability, bias assessment, uncertainty quantification, evidence quality hierarchy

Collection Statistics

Metric Value
Queries investigated 3
High confidence (all four features absent in journalism) 1 (Q001)
High confidence (all disciplines are formal) 1 (Q002)
High confidence (taxonomy remains conceptual) 1 (Q003)

Source Independence Assessment

Sources across the three queries are highly independent. Q001 draws on journalism organizations (IFCN, PolitiFact, NewsGuard, BBC, Bellingcat, SPJ, EJC). Q002 draws on cross-disciplinary standards (legal, auditing, epidemiology, engineering, historical, information science). Q003 draws on policy and training sources (Council of Europe, First Draft, EU, Verification Handbook). The only significant overlap is the Verification Handbook, which appears in both Q001 (as a journalism framework) and Q003 (as a venue for taxonomy adoption). This overlap is appropriate — the handbook legitimately addresses both topics.

Collection Gaps

Gap Impact Mitigation
Internal newsroom verification procedures (unpublished) Medium — may contain structural features not in public standards Would require newsroom access; not available through open-source research
Platform-internal content moderation procedures Medium — may operationalize Wardle-Derakhshan taxonomy internally Would require platform disclosure; not publicly available
Non-English journalism and fact-checking frameworks Low — may contain structural features not in English-language literature Broader search needed for comprehensive coverage
Forensic science and actuarial evidence standards Low — additional disciplines for Q002 completeness Could extend the survey but unlikely to change key findings

Collection Self-Audit

Domain Rating Notes
Eligibility criteria Low risk All three queries had clear, predefined criteria
Search comprehensiveness Low risk 11 search rounds across three queries; 220+ results dispositioned
Evaluation consistency Low risk Same scorecard format and hypothesis structure across all queries
Synthesis fairness Low risk Contradictory evidence surfaced in all three queries; three hypotheses tested for each

Resources

Summary

Metric Value
Queries investigated 3
Files produced 134
Sources scored 22
Evidence extracts 22
Results dispositioned 28 selected + 192 rejected = 220 total

Tool Breakdown

Tool Uses Purpose
WebSearch 16 Search queries across all three investigations
WebFetch 8 Page content retrieval for detailed extraction
Write 45 File creation for output archive
Read 2 Methodology prompt and output format spec
Edit 0 No edits needed
Bash 12 Directory creation, batch file generation

Token Distribution

Category Tokens
Input (context) ~200,000
Output (generation) ~80,000
Total ~280,000