DR-301d · Module 1
Source Classification Taxonomy
3 min read
Sources fall into four classes based on their relationship to the information they provide. Primary sources generated the information — SEC filings, patent applications, earnings transcripts, company press releases. They are closest to the truth but may have self-serving framing. Secondary sources reported on the information — analyst reports, news coverage, industry publications. They add interpretation but introduce editorial bias. Tertiary sources aggregated from secondary sources — Wikipedia summaries, market overview slideshows, general business articles. They are convenient but often inaccurate on specifics. Derived sources used the information as input to their own analysis — competitor benchmarks, market sizing models, trend reports. They add value through synthesis but compound the errors of their inputs.
Do This
- Trace every finding back to its primary source — even if you found it in a secondary source
- Label every source by class in your research artifacts — the reader should know the provenance
- Weight primary sources above secondary in any conflict between them
- Use tertiary sources for orientation only — never as evidence in a brief
Avoid This
- Treat an analyst report as a primary source — it is secondary, with the analyst's interpretation layered on
- Accept a secondary source's framing without checking the primary source it cites
- Build conclusions on tertiary sources — they are summaries of summaries
- Ignore source class when two sources conflict — the primary source wins by default