1254 Source Evaluation 9bc285fb

Source Evaluation Criteria

Detailed criteria for assessing source reliability and authority.

The CRAAP Test (Modified for AI Research)

For each source, evaluate:

Currency

Red flags:

• No date anywhere on the page
• References to "new" features that are now standard
• Broken external links
• Deprecated syntax or APIs

Relevance

Authority

Authority indicators by domain:

Accuracy

Verification techniques:

1. Check if cited sources actually support the claims
2. Cross-reference with official documentation
3. Test code examples if possible
4. Look for corroborating independent sources

Purpose

Bias indicators:

• Affiliate links without disclosure
• Only positive coverage of a product
• Attacks on competitors
• Sponsored content not clearly labeled
• Author works for company being discussed

Source Type Reliability Hierarchy

From most to least reliable (general guidelines, exceptions exist):

1. Primary sources — Official documentation, original research, firsthand accounts
2. Academic/peer-reviewed — Journal articles, conference papers, institutional research
3. Authoritative secondary — Well-sourced news from reputable outlets, industry analyst reports
4. Expert analysis — Known experts writing in their field, with citations
5. Community consensus — High-vote Stack Overflow answers, widely-used GitHub repos
6. General secondary — Wikipedia (follow its sources), established tech blogs
7. Individual opinions — Personal blogs, forum posts, social media
8. Unknown/anonymous — Unsourced claims, SEO content farms, AI-generated without verification

Red Flags That Lower Source Trust

Content Red Flags

• No author attribution
• No publication or update date
• No citations or references
• Claims that can't be verified
• Sensational or clickbait headlines
• Grammatical errors or poor writing quality
• Contradicts well-established facts
• "According to experts" without naming them

Technical Red Flags

• Deprecated or non-working code examples
• Incorrect technical terminology
• Advice that contradicts official docs
• Missing security considerations
• Platform/version mismatch with current reality

Structural Red Flags

• Excessive ads or pop-ups
• Content hidden behind paywalls with no credentials
• Domain looks like a content farm
• No About page or organizational info
• Recently created domain with authoritative claims

Evaluating AI-Generated Content

AI content is increasingly common; evaluate carefully:

Warning signs of unverified AI content:

• Generic, surface-level coverage
• Confident claims without citations
• Plausible-sounding but incorrect technical details
• Inconsistencies within the same article
• Lack of specific examples or real-world context

Mitigation:

• Never trust AI summaries as primary sources
• Always trace claims back to original sources
• Verify technical claims against official docs
• Test code examples before accepting them

Source Evaluation Checklist

Before citing any source, verify:

- [ ] Author/organization identified and credible
- [ ] Publication date is acceptable for topic
- [ ] Claims are consistent with other sources
- [ ] Technical content is current and accurate
- [ ] No obvious commercial bias
- [ ] URL is functional and content is accessible

Recording Source Quality

When documenting research, include:

| Source | Type | Authority | Recency | Bias Risk | Notes |
|--------|------|-----------|---------|-----------|-------|
| [URL] | Official docs | High | Current | Low | Primary source |
| [URL] | Blog post | Medium | 2023 | Medium | Author works at competitor |
| [URL] | Research paper | High | 2022 | Low | Peer-reviewed, highly cited |