Grey literature covers government reports, regulatory submissions, theses, conference proceedings and technical documents from institutions.
Grey literature covers government reports, regulatory submissions, theses, conference proceedings and technical documents from institutions. It is produced to be used rather than to be sold, and it is frequently where the most detailed data sit.
Documents submitted to regulators often contain full trial data, including studies never published anywhere. Several important reassessments of widely used treatments came from reading those submissions rather than the journal literature.
Because grey literature is not filtered by a journal's appetite for novelty, it contains findings that would otherwise be invisible. Reviews searching it find systematically different pictures from reviews that do not.
Without a uniform review process, quality ranges from rigorous to promotional. The absence of a gatekeeper cuts both ways, and the burden of assessment moves entirely onto the reader.
It is scattered across agency sites, repositories and archives, and indexed inconsistently. This is the practical reason reviews skip it, and skipping it is a known source of bias rather than a neutral choice.
Doctoral theses contain full methods and complete results, including the analyses that did not make the resulting paper. They are deposited publicly by most institutions and are rarely consulted.
Name what the document is, who produced it, and when, because a reader cannot assume a review process. Treating a report as equivalent to a reviewed paper, or dismissing it as not real research, are both wrong.
A synthesis limited to journals is a synthesis of what journals chose to publish. For any question where publication bias is plausible, that limitation shapes the conclusion more than the analysis does.
Systematic comparisons find that reviews including unpublished and regulatory sources report smaller effects than those limited to journals. The difference is the publication filter made visible, and it is one of the better-documented findings about how evidence synthesis goes wrong.
Regulator sites, institutional repositories and thesis collections are indexed separately from journal databases and often poorly by general search. Searching each directly is slower than one database query and is where the material that changes a conclusion tends to sit.
The claims on this page about reports, theses and regulatory documents that never reach a journal can be checked against sources outside this site: