Secondary source

A secondary source is a publication that draws on existing research rather than reporting findings from a study the authors conducted themselves. Reviews, textbooks, database entries, clinical guidelines, and community-maintained compilations all qualify. Their authors synthesise, summarise, or comment on primary literature — the original studies, trials, and reports in which data were first collected and analysed.

In the evidence hierarchy a reader encounters on these pages, secondary sources are typically the most accessible layer. A claim in a review article, a pharmacology reference, or a harm-reduction database almost always traces back to one or more primary sources — the peer-reviewed papers, case reports, or clinical trials where the fact was first established.

Psychedex records source type as part of provenance. Where a value rests only on secondary literature, that is noted, because traceability is not the same when the citation points to a summary rather than to an original measurement.

What this design can establish

A well-sourced secondary publication can map what the research community at a given time understood about a topic — the positions that had accumulated enough primary-literature support to appear in reviews and reference works. Broad agreement across multiple independent secondary sources, each citing distinct primary papers, is reasonable evidence that a claim appears in the underlying research record.

Secondary sources are also the practical means by which specialists in one field track findings from another. Their value lies in synthesis: condensing a literature that no individual reader has time to follow in full.

What it cannot

A secondary source cannot establish the accuracy or quality of the data it reports. An error introduced in a primary paper, or a finding taken out of context, can propagate through dozens of review articles and database entries without any of them catching the problem.

Many widely cited secondary-source claims trace back, when followed to their root, to a single study — sometimes a small one, sometimes one whose methodological limits the citing reviews do not mention. The inference a reader is most likely to make — that a claim appearing across many sources has been independently confirmed many times — is precisely the one to resist. Repetition and verification are not the same thing.

The clearest case is an absent primary record. If no original research has been published on a specific question, that gap will appear in secondary sources but cannot be closed by them. A secondary source that states 'no data available' is accurate; one that fills the gap with a plausible-sounding estimate converts an absence into an apparent fact without a basis for it.