Wikipedia:Evaluating sources

This essay examines how to evaluate sources within the context of Wikipedia's content policies. Part of evaluating a source is deciding whether it is a primary, secondary, or tertiary source. For the policy on sourcing issues, see this section of No original research, and Verifiability. If there are inconsistencies between this page and the policies, please update this page.

The decision as to which sources are appropriate in any given situation is a matter of common sense and good editorial judgment, and should be discussed on individual article talk pages to achieve consensus. In cases where a consensus is not forthcoming, it may be helpful to seek some assistance in reaching an agreement.

Types of sources
Sources of information are commonly categorized as primary, secondary, or tertiary sources. In brief, a primary source is one close to the event with firsthand knowledge (for example, an eyewitness); a secondary source is at least one step removed (for example, a book about an event written by someone not involved in it); and a tertiary source is an encyclopaedia or textbook that provides a general overview.

The way these concepts are applied to particular sources can change over time. A newspaper article that we regard as a secondary source now, might be regarded as a primary source in 100 years, because it would be close to the event in relation to those reading about that event 100 years from now.

Primary, secondary, and tertiary sources

 * Primary sources are sources of original work, as well as historical items and references close to the subject. Depending on the field, this can range from speeches, personal correspondence, published editorials, manuscripts, works of fiction, incidents captured on film, photographs, artwork, witness reports, legal documents, laboratory notebooks, field notes, peer-reviewed articles publishing original research, and even artifacts.


 * Secondary sources are reports that draw on research and other references to make interpretive, analytical, or synthesized claims. Depending on the field, these may include textbooks, review articles, and peer-reviewed articles publishing original research. They are best used for representing significant points of view.


 * Tertiary sources are materials that provide an overview of primary and secondary sources, such as encyclopedias, textbooks, and other compendia. Wikipedia is a tertiary source.

These definitions are not mutually exclusive. Primary sources, for example, might draw on secondary sources to make interpretive, analytic, or synthetic claims. In such cases, sources should be evaluated according to how they are used.

Examples by field
Table source: Saylor, Ward & Hooper, Helen – James Cook University

Original research and verifiability
No Original Research and Verifiability are the two core content policies that determine how sources are used on Wikipedia. All sources should be used in a way that does not give rise to analyses, syntheses, or original conclusions. Original analysis by Wikipedians may not be added to articles.

Edits should not rely on unclear or inconsistent passages. Passages should not be taken out of context in a way that changes their meaning.

Primary sources that have been published by a reliable source may be used, but any part of an article that relies on a primary source should:
 * only make descriptive claims about the material found in the primary source, the accuracy and applicability of which is easily verifiable by any reasonable, educated person without specialist knowledge, and
 * make no analytic, synthetic, interpretive, explanatory, or evaluative claims about the information found in the primary source.

Tertiary sources can be useful in providing context and avoiding original research in topics where there exist very large amounts of primary or secondary sources.

Biographies of living persons
Material about living persons must be sourced very carefully. Without reliable secondary sources, it will violate the No original research, Verifiability, and Notability policies, as well as Biographies of living persons.

Material about living persons must be carefully evaluated to ensure neutral point of view, and avoid undue weight. The use of primary sources may easily lead to syntheses that constitute original research. Editors should not use public records as a source about a subject &mdash; such as birth certificates, home evaluations, traffic citations, vehicle registrations, or trial transcripts &mdash; unless these have been used by a reliable, secondary source. For example, if writing about a subject who had a messy divorce, do not go the courthouse to retrieve court papers about it; instead, rely on secondary sources, such as mainstream newspapers, that have written about the divorce. If they have not written about it, nor should Wikipedia.

Notability
A subject merits its own article on Wikipedia when it is notable enough to have received acknowledgment in multiple, reliable, secondary sources. If an article lacks secondary sources that reasonably demonstrate its importance, it may be listed for deletion.

In the sciences
When evaluating sources of scientific material, sources that were evaluated by someone other than the scientist / author when they were published are the most reliable. This includes peer-reviewed articles in scientific journals. The general rule is to always consider the source. The critical element in evaluating the source of material that is to be included in Wikipedia was previously published in a source that reviewed that information. Journal articles that have passed through the peer-review process are thus the most highly reliable sources of scientific information.
 * Peer-reviewed literature

Peer-reviewed literature is reliable as a source for propositions, ideas, observations, and other scientific data. However, care must be taken that multiple propositions, ideas, observations, and other data are not stitched together to form a new proposition or idea. The sum total of a piece of writing in Wikipedia should not advance a unique or novel interpretation. This is original research, even if each individual component is sourced to highly reliable peer-reviewed literature; original research may not be published in Wikipedia.

Newspaper articles, encyclopedias, and textbooks that distill peer-reviewed literature and scientific findings into lay-person language may be easier to understand for the non-expert. If available, they must be used carefully as they are further interpretations of the original work. This sort of coverage is not available for all scientific topics, and is not required to establish reliability of a scientific proposition. Lay-person language interpretations, if confusing or inaccurate, should be buttressed by reference to the original material. In a conflict, the peer-reviewed publication is more authoritative and reliable.
 * Newspaper articles, encyclopedias, textbooks, and other non-peer-reviewed sources

Scientific findings that are originally presented in non-peer reviewed literature must be used with particular care, as they have not been reviewed for scientific accuracy. For example, a newspaper may run an article on a scientist and her latest work, or a scientist might maintain a blog about his research. The descriptions of the scientist's previously unpublished findings, as published in the newspaper or on the blog, are not as reliable a description of that research as a peer-reviewed journal article. If it is important to discuss such findings in the Wikipedia article, the most reliable source available should be used. If a peer-reviewed article is published after a newspaper, blog, or other non-peer-reviewed publication of the research, both may be used, but in a conflict, the peer-reviewed publication is more authoritative and reliable.

Review articles are articles published within scientific journals that survey and synthesize the state of research in a particular area. Review articles can be very helpful in understanding a topic, and on-point review articles should be cited or included as "further research". Review articles, like encyclopedias or textbooks, may also be useful to cite for general propositions about a field. However, there are three issues to consider when using review articles. First, like any restated material, review articles may have errors. In a conflict over what a paper said, the peer-reviewed publication is more authoritative and reliable than a review article's summation of the publication's findings. Second, a review article may summarize later research or findings that shed new light on earlier research. If possible, the summarized later research should be reviewed and cited directly when describing any points from that research; however, it is permissible to reference the review article as citing the earlier material. Third, a review article that advances new information or its own new synthesis may be cited for those propositions, just as any publication may. However, the source should be carefully evaluated because review articles may not be peer reviewed
 * Review articles