source-credibility

Assess the credibility of $ARGUMENTS.

Step 1: Identify the source type

Different source types have different credibility frameworks:

Type	What to assess
News publication	Ownership, editorial standards, corrections policy, press council membership
Trade/industry publication	Who funds it, advertiser relationships, editorial independence
Think tank / research organisation	Funding sources, declared mission, publication peer review
Government / regulatory body	Statutory mandate, transparency obligations
Academic institution	Peer review process, retraction history, funding disclosure
Company / PR	By definition advocacy — assess transparency, not neutrality
Individual / expert	Credentials, institutional affiliation, track record, conflicts of interest
Aggregator / social platform	No editorial standard — assess the underlying sources it links to

Step 2: Ownership and funding

Who owns the source and who funds it? This doesn't determine whether individual pieces are accurate, but it shapes systematic biases and blind spots.

Search the source's "About" page, press releases, and public records. For significant publications:

• Media ownership databases: MEAA (AU media), Muck Rack, Columbia Journalism Review
• Company registration: ASIC Connect (AU), Companies House (UK) for ownership chains
• Known AU media ownership: News Corp AU (The Australian, Herald Sun, Daily Telegraph), Nine Entertainment (SMH, The Age, AFR), Seven West Media (The West Australian)
• NZ media ownership: NZME (NZHerald), Stuff

For think tanks and research organisations: check annual reports, funding pages, and donation transparency disclosures.

Step 3: Editorial standards

Does the source have published editorial standards? Assess:

• Corrections policy: does it publish corrections? How prominently?
• Press council or regulatory membership: AU — ACMA for broadcast, Press Council for print/digital; NZ — Broadcasting Standards Authority, NZ Press Council
• Bylines and accountability: are authors named and contactable?
• Source attribution standards: does it name sources or rely heavily on anonymous attribution?

Publications outside a press council or regulatory body have no external accountability mechanism.

Step 4: Track record

Has the source been accurate and reliable historically?

• Search for notable corrections, retractions, or failures
• Check Media Bias / Fact Check — useful as a starting point, though it has its own limitations; use alongside other sources
• Look for: fabricated stories, misrepresentation of studies, consistent pattern of one-sided coverage on specific topics
• For academic sources: retraction databases (Retraction Watch, PubMed retraction notices)

A single error doesn't define a source. A pattern does.

Step 5: Declared mission and known biases

What does the source say its purpose is? Is that consistent with its output?

Most sources have a perspective — the question is whether they're transparent about it. A declared advocacy organisation that publishes transparently is more credible than a neutral-seeming publication with undisclosed funding.

Common bias patterns to identify:

• Selection bias: consistently covers certain topics, ignores others
• Framing bias: systematically frames issues in particular directions
• Source bias: routinely quotes from a narrow range of perspectives
• Commercial bias: coverage aligns with advertiser or funder interests

Note the distinction between bias (a systematic pattern) and error (a specific inaccuracy). Both matter but they're different problems.

Step 6: For individual experts

Assess:

• Credentials: does the claimed expertise match the topic? (A cardiologist commenting on climate science is outside their domain)
• Institutional affiliation: current role, past roles
• Funding and conflicts: has the expert received funding from parties with a stake in their conclusions?
• Track record: have their claims in this domain held up over time?
• Citation and peer engagement: are they cited by other credentialed people in their field, or primarily by media?

Rules

• Credibility is not binary. Assess on a spectrum and be specific about which dimensions are strong or weak.
• Bias ≠ inaccuracy. A biased source can report individual facts accurately. A neutral-seeming source can publish errors.
• Funding creates incentives, not conclusions. Note it as a flag, not proof of corruption.
• No source is perfectly credible. The question is whether the credibility is sufficient for the intended use.
• Absence of information about a source is itself a credibility signal — opacity is a flag.

Output format

## Source credibility: [Source name]

**Date of assessment:** [today]
**Source type:** [news / trade / think tank / government / academic / company / individual / other]

### Ownership and funding

[Who owns it, who funds it, transparency of disclosure]

### Editorial standards

[Press council membership, corrections policy, bylines, attribution standards]

### Track record

[Notable accuracy failures or strong reliability record — specific examples where found]

### Declared mission and known biases

[What the source says it is vs what the output pattern shows]

### Credibility assessment

| Dimension | Rating | Evidence |
|---|---|---|
| Ownership transparency | High / Medium / Low / Unknown | — |
| Editorial accountability | High / Medium / Low / Unknown | — |
| Accuracy track record | High / Medium / Low / Unknown | — |
| Bias transparency | High / Medium / Low / Unknown | — |

**Overall credibility:** [High / Medium / Low / Insufficient information]

**Appropriate use:** [What this source is reliable for, and what it's not]

### Sources used in this assessment

1. [Source](URL)