This page in a nutshell: Each book, article, or other source needs to be credible for supporting the particular claim(s) it supports. In case of a dispute, editors need to be able to explain why they relied on that source.
Wikipedia articles should be based on reliable, published sources, making sure that all majority and significant minority views that have appeared in those sources are covered (see Wikipedia:Neutral point of view). If no reliable sources can be found on a topic, Wikipedia should not have an article on it.
This guideline discusses the reliability of various types of sources. The policy on sourcing is Wikipedia:Verifiability, which requires inline citations for any material challenged or likely to be challenged, and for all quotations. The verifiability policy is strictly applied to all material in the mainspace—articles, lists, and sections of articles—without exception, and in particular to biographies of living persons, which states:
Contentious material about living persons (or, in some cases, recently deceased) that is unsourced or poorly sourced—whether the material is negative, positive, neutral, or just questionable—must be removed immediately and without waiting for discussion.
Source reliability falls on a spectrum: No source is 'always reliable' or 'always unreliable' for everything. However, some sources provide stronger or weaker support for a given statement. Editors must use their judgment to draw the line between usable and inappropriate sources for each statement.
Articles should be based on reliable, independent, published sources with a reputation for fact-checking and accuracy. This means that we publish only the analysis, views, and opinions of reliable authors, and not those of Wikipedians, who have read and interpreted primary source material for themselves. The following examples cover only some of the possible types of reliable sources and source reliability issues, and are not intended to be exhaustive. Proper sourcing always depends on context; common sense and editorial judgment are an indispensable part of the process.
A source is where the material comes from. For example, a source could be a book or a webpage. A source can be reliable or unreliable for the material it is meant to support. Some sources, such as unpublished texts and an editor's own personal experience, are prohibited.
When editors talk about sources that are being cited on Wikipedia, they might be referring to any one of these three concepts:
Any of the three can affect reliability. Reliable sources may be published materials with a reliable publication process, authors who are regarded as authoritative in relation to the subject, or both. These qualifications should be demonstrable to other people.
Definition of published
"WP:PUBLISHED" redirects here. Not to be confused with
Published means, for Wikipedia's purposes, any source that was made available to the public in some form. The term is most commonly associated with text materials, either in traditional printed format or online; however, audio, video, and multimedia materials that have been recorded then broadcast, distributed, or archived by a reputable party may also meet the necessary criteria to be considered reliable sources. Like text, media must be produced by a reliable source and be properly cited. Additionally, an archived copy of the media must exist. It is convenient, but by no means necessary, for the archived copy to be accessible via the Internet.
The reliability of a source depends on context. Each source must be carefully weighed to judge whether it is reliable for the statement being made in the Wikipedia article and is an appropriate source for that content.
In general, the more people engaged in checking facts, analyzing legal issues, and scrutinizing the writing, the more reliable the publication. Sources should directly support the information as it is presented in the Wikipedia article.
Information provided in passing by an otherwise reliable source or information that is not related to the principal topics of the publication may not be reliable; editors should cite sources focused on the topic at hand where possible. For example, a publisher's web site is likely to be reliable for an author's identity, date of publication, etc., but not necessarily for a critical, artistic, or commercial evaluation of the work (see § Reliability in specific contexts, below).
Especially in scientific and academic fields, older sources may be inaccurate because new information has been brought to light, new theories proposed, or vocabulary changed. In areas like politics or fashion, laws or trends may make older claims incorrect. Be sure to check that older sources have not been superseded, especially if it is likely that new discoveries or developments have occurred in the last few years. In particular, newer sources are generally preferred in medicine.
Sometimes sources are too new to use, such as with breaking news (where later reports might be more accurate), and primary sources which purport to debunk a long-standing consensus or introduce a new discovery (in which case awaiting studies that attempt to replicate the discovery might be a good idea, or reviews that validate the methods used to make the discovery).
Similarly for breaking news, a contemporary secondary news source can quickly become a historical primary source. Articles of recent current events must be periodically updated with new secondary sources.
Sources of any age may be prone to recentism, and this needs to be balanced out by careful editing.
How accepted and high-quality reliable sources use a given source provides evidence, positive or negative, for its reliability and reputation. The more widespread and consistent this use is, the stronger the evidence. For example, widespread citation without comment for facts is evidence of a source's reputation and reliability for similar facts, whereas widespread doubts about reliability weigh against it. If outside citation is the main indicator of reliability, particular care should be taken to adhere to other guidelines and policies, and to not unduly represent contentious or minority claims. The goal is to reflect established views of a topic as far as we can determine them.
Many Wikipedia articles rely on scholarly material. When available, academic and peer-reviewed publications, scholarly monographs, and textbooks are usually the most reliable sources. However, some scholarly material may be outdated, in competition with alternative theories, controversial within the relevant field, or largely ignored by the mainstream academic discourse because of lack of citations. Try to cite current scholarly consensus when available, recognizing that this is often absent. Reliable non-academic sources may also be used in articles about scholarly issues, particularly material from high-quality mainstream publications. Deciding which sources are appropriate depends on context. Material should be
Prefer secondary sources – Articles should rely on secondary sources whenever possible. For example, a paper reviewing existing research, a review article, monograph, or textbook is often better than a primary research paper. When relying on primary sources, extreme caution is advised. Wikipedians should never interpret the content of primary sources for themselves (see Wikipedia:No original research and Wikipedia:Neutral point of view).
Reliable scholarship – Material such as an article, book, monograph, or research paper that has been vetted by the scholarly community is regarded as reliable, where the material has been published in reputable peer-reviewed sources or by well-regarded academic presses.
Dissertations – Completed dissertations or theses written as part of the requirements for a doctorate, and which are publicly available (most via interlibrary loan or from ProQuest), can be used but care should be exercised, as they are often, in part, primary sources. Some of them will have gone through a process of academic peer reviewing, of varying levels of rigor, but some will not. If possible, use theses that have been cited in the literature; supervised by recognized specialists in the field; or reviewed by independent parties. Dissertations in progress have not been vetted and are not regarded as published and are thus not reliable sources as a rule. Some theses are later published in the form of scholarly monographs or peer reviewed articles, and, if available, these are usually preferable to the original thesis as sources. Masters dissertations and theses are considered reliable only if they can be shown to have had significant scholarly influence.
Citation counts – One may be able to confirm that discussion of the source has entered mainstream academic discourse by checking what scholarly citations it has received in
themselves
often.
Isolated studies – Isolated studies are usually considered tentative and may change in the light of further academic research. If the isolated study is a primary source, it should generally not be used if there are secondary sources that cover the same content. The reliability of a single study depends on the field. Avoid
review articles
are preferred when available, so as to provide proper context.
POV and peer review in journals – Care should be taken with journals that exist mainly to promote a particular point of view. A claim of peer review is not an indication that the journal is respected, or that any meaningful peer review occurs. Journals that are not peer reviewed by the wider academic community should not be considered reliable, except to show the views of the groups represented by those journals.[note 1]
self-published sources. These include preprints, which are preliminary versions of articles intended for later publication in peer-reviewed journals, as well as other materials like lecture notes and informal expository writing. These materials are seldom reliable sources, and they are not permitted for medical content
. However, they sometimes meet the criteria for acceptable use of self-published sources, and links to such repositories can be used as open-access links for papers which have subsequently been published reliably.
News sources often contain both factual content and opinion content. News reporting from well-established news outlets is generally considered to be reliable for statements of fact (though even the most reputable reporting sometimes contains errors). News reporting from less-established outlets is generally considered less reliable for statements of fact. Most newspapers also reprint items from news agencies such as Reuters, Interfax, Agence France-Presse, United Press International or the Associated Press, which are responsible for accuracy. The agency should be cited in addition to the newspaper that reprinted it.
Scholarly sources and high-quality non-scholarly sources are generally better than news reports for academic topics (see § Scholarship, above). Press releases from organizations or journals are often used by newspapers with minimal change; such sources are churnalism and should not be treated differently than the underlying press release. Occasionally, some newspapers still have specialist reporters who are citable by name. (For topics relating to health or medicine, see § Medical claims, below.)
Otherwise reliable news sources—for example, the website of a major news organization—that publish in a blog-style format for some or all of their content may be as reliable as if published in standard news article format (See also Wikipedia:Verifiability § Newspaper and magazine blogs).
Signals that a news organization engages in fact-checking and has a reputation for accuracy are the publication of corrections and disclosures of conflicts of interest.
Human interest reporting is generally not as reliable as news reporting, and may not be subject to the same rigorous standards of fact-checking and accuracy (see Junk food news).[6]
The reporting of rumors has a limited encyclopedic value, although in some instances verifiable information about rumors may be appropriate (i.e. if the rumors themselves are noteworthy, regardless of whether or not they are true).
Wikipedia is not the place for passing along gossip and rumors
.
Some news organizations have used Wikipedia articles as a source for their work. Editors should therefore beware of
statements attributed to that editor or author, but are rarely reliable for statements of fact (see also § Statements of opinion
, below).
When taking information from opinion content, the identity of the author may help determine reliability. The opinions of specialists and recognized experts are more likely to be reliable and to reflect a significant viewpoint.[note 4] If the statement is not authoritative, attribute the opinion to the author in the text of the article and do not represent it as fact. Reviews for books, movies, art, etc. can be opinion, summary, or scholarly pieces.[7][8]
Some news organizations may not publish their editorial policies.
Some websites function partly or entirely as aggregators, reprinting items from websites of news agencies, blogs, websites, or even Wikipedia itself. These may constitute a curated feed or an AI-generated feed. Examples include the main pages of MSN and Yahoo News. As with newspaper reprints, the original content creator is responsible for accuracy and reliability should be judged based on the original source. Direct links to the original source should be preferred over the aggregator's link.
such things as titles and running times. Journalistic and academic sources are preferable, however, and e-commerce links should be replaced with reliable non-commercial sources if available.
Rankings proposed by vendors (such as bestseller lists at Amazon) usually have at least one of the following problems:
It may be impossible to provide a stable source for the alleged ranking.
When only self-published by the vendor, i.e. no reliable independent source confirming the ranking as being relevant, the ranking would usually carry insufficient weight to be mentioned in any article.
For such reasons, such rankings are usually avoided as Wikipedia content.
Wikipedia articles are required to present a neutral point of view. However, reliable sources are not required to be neutral, unbiased, or objective. Sometimes non-neutral sources are the best possible sources for supporting information about the different viewpoints held on a subject.
Common sources of bias include political, financial, religious, philosophical, or other beliefs. Although a source may be biased, it may be reliable in the specific
in-text attribution appropriate, as in "The feminist Betty Friedan wrote that..."; "According to the Marxist economist Harry Magdoff..."; or "The conservative Republican presidential candidate Barry Goldwater
Questionable sources are those with a poor reputation for checking the facts or with no editorial oversight. Such sources include websites and publications expressing views that are widely acknowledged as extremist, that are promotional in nature, or that rely heavily on rumors and personal opinions.[9] Questionable sources are generally unsuitable for citing contentious claims about third parties, which includes claims against institutions, persons living or dead, as well as more ill-defined entities. The proper uses of a questionable source are very limited.
Beware of sources that sound reliable but do not have the
reputation for fact-checking and accuracy that this guideline requires.[10] The Journal of 100% Reliable Factual Information might have a reputation for "predatory" behavior, which includes questionable business practices and/or peer-review processes that raise concerns about the reliability of their journal articles.[11][12]
primary sources in articles. Reliable publications clearly indicate sponsored articles in the byline or with a disclaimer
at the top of the article. Sources that do not clearly distinguish staff-written articles from sponsored content are also questionable.
Symposia and supplements to academic journals are often (but far from always) unacceptable sources. They are commonly sponsored by industry groups with a financial interest in the outcome of the research reported. They may lack independent editorial oversight and peer review, with no supervision of content by the parent journal.[13] Such articles do not share the reliability of their parent journal,[14] being essentially paid ads disguised as academic articles. Such supplements, and those that do not clearly declare their editorial policy and conflicts of interest, should not be cited.
Indications that an article was published in a supplement may be fairly subtle; for instance, a letter "S" added to a page number,
The Times Higher Education Supplement
. A sponsored supplement also does not necessarily involve a COI; for instance, public health agencies may also sponsor supplements. However, groups that do have a COI may hide behind layers of front organizations with innocuous names, so the ultimate funding sources should always be ascertained.
Examples of unacceptable user-generated sources are
self referencing
). For official accounts from celebrities and organizations on social media, see the section about self-published sources below.
Although review aggregators (such as Rotten Tomatoes and Metacritic) may be reliable when summarizing experts, the ratings and opinions of their users (including the reported rating averages) are not.
In particular, a wikilink is not a reliable source.
Self-published and questionable sources as sources on themselves
Self-published or questionable sources may be used as sources of information about themselves, especially in articles about themselves, without the requirement that they be published experts in the field, so long as the following criteria are met:
The material is neither unduly self-serving nor an exceptional claim.
It does not involve claims about third parties (such as people, organizations, or other entities).
It does not involve claims about events not directly related to the subject.
There is no reasonable doubt as to its authenticity.
The Wikipedia article is not based primarily on such sources.
These requirements also apply to pages from social networking websites such as Twitter, Tumblr, and Facebook. Use of self-sourced material should be minimal; the great majority of any article must be drawn from independent sources.
Since the AI boom in the early 2020s, so-called generative artificial intelligence ("generative AI" or "GenAI"), specifically large language models (LLMs), have become a common tool to generate and publish material using machine learning (ML). It may not be known or detectable that ML was used to produce a given piece of text.
Content from LLM tools such as
AI chatbots is not generally reliable. LLMs are trained using text scraped from the internet (including Wikipedia) and suffer many of the same problems as self-published and user-generated content (see § User-generated content
, above).
LLMs also have a tendency to "hallucinate" false information, including source citations that look as if they are from reputable publications but do not actually exist.[note 5] Such spurious material may be generated unintentionally by writers, reporters, scientists, medical researchers, and lawyers using chatbots to help them to produce reports, or maliciously by individuals and companies publishing "fake news".
Editors must take particular care when writing biographical material about living persons. Contentious material about a living person that is unsourced or poorly sourced should be removed immediately; do not move it to the talk page. This applies to any material related to living persons on any page in anynamespace, not just article space.
Wikipedia articles should be based mainly on reliable
secondary sources
, i.e., a document or recording that relates to or discusses information originally presented elsewhere.
Reputable
WP:SELFSOURCE
).
Primary sources are often difficult to use appropriately. Although they can be both reliable and useful in certain situations, they must be used with caution in order to avoid original research
. Although specific facts may be taken from primary sources, secondary sources that present the same material are preferred. Large blocks of material based purely on primary sources should be avoided. All interpretive claims, analyses, or synthetic claims about primary sources must be referenced to a secondary source, rather than original analysis of the primary-source material by Wikipedia editors.
When editing articles in which the use of primary sources is a concern, in-line templates, such as {{
and position statements from nationally or internationally reputable expert bodies. It is vital that the biomedical information in all types of articles be based on reliable, independent, published sources and accurately reflect current medical knowledge.
Inclusion and exclusion of content related to fringe theories and criticism of fringe theories may be done by means of a rough parity of sources. If an article is written about a well-known topic about which many peer-reviewed articles are written, it should not include fringe theories that may seem relevant but are only sourced to obscure texts that lack peer review. Parity of sources may mean that certain fringe theories are only reliably and verifiably reported on, or criticized, in alternative venues from those that are typically considered reliable sources for scientific topics on Wikipedia.
In an article on a fringe topic, if a notable fringe theory is primarily described by amateurs and self-published texts, verifiable and reliable criticism of the fringe theory need not be published in a peer-reviewed journal. For example, the
biographies of living persons
policies are not suspended simply because the topic is a fringe theory.
The accuracy of quoted material is paramount and the accuracy of quotations from living persons is especially sensitive. To ensure accuracy, the text of quoted material is best taken from (and cited to) the original source being quoted. If this is not possible, then the text may be taken from a reliable secondary source (ideally one that includes a citation to the original). No matter where you take the quoted text from, it is important to
make clear the actual source of the text, as it appears in the article
.
Partisan secondary sources should be viewed with suspicion as they may misquote or quote out of context. In such cases, look for neutral corroboration from another source.
Any analysis or interpretation of the quoted material, however, should rely on a secondary source (see Wikipedia:No original research).
A statement that all or most scientists or scholars hold a certain view requires reliable sourcing that directly says that all or most scientists or scholars hold that view. Otherwise, individual opinions should be identified as those of particular, named sources. Editors should avoid original research especially with regard to making blanket statements based on
novel syntheses of disparate material. Stated simply, any statement in Wikipedia that academic consensus exists on a topic must be sourced rather than being based on the opinion or assessment of editors. Review articles
, especially those printed in academic review journals that survey the literature, can help clarify academic consensus.
Some sources may be considered reliable for statements as to their author's opinion, but not for statements asserted as fact. For example, an inline qualifier might say "[Author XYZ] says....". A prime example of this is opinion pieces in mainstream newspapers. When using them, it is best to clearly attribute the opinions in the text to the author and make it clear to the readers that they are reading an opinion (see also § Editorial and opinion commentary, above).
Breaking-news reports often contain serious inaccuracies. As an electronic publication, Wikipedia can and should be up to date, but
Wikipedia is not a newspaper and it does not need to go into all details of a current event in real time. It is better to wait a day or two after an event before adding details to the encyclopedia, than to help spread potentially false rumors. This gives journalists time to collect more information and verify claims, and for investigative authorities to make official announcements. The On the MediaBreaking News Consumer's Handbook[19]
contains several suggestions to avoid spreading unreliable and false information. These include: distrust anonymous sources, unconfirmed reports, and reports attributed to other news media; seek multiple independent sources which independently verify; seek verified eyewitness reports; and be wary of potential hoaxes. With mass shootings, remain skeptical of early reports of additional attackers, coordinated plans, and bomb threats.
For health- and science-related breaking-news, Wikipedia has specific sourcing standards to prevent inaccuracies: see
press releases), to prefer reports which avoid hyperbolic language and describe both benefits and costs of a new treatment (all treatments have trade-offs), to be wary of disease mongering (exaggerating risks, symptoms, or anecdotes of a disease which leads to unnecessary worry, panic, or spending), and to be skeptical of treatments which are "awaiting FDA approval" or in pre-clinical testing" as more than 90% of all treatments fail during these stages and,[21] even if efficacious, may be 10 to 15 years or more from reaching the consumer market.[22]
—are not a reliable source. If the information is supported by the body of the source, then cite it from the body. Headlines are written to grab readers' attention quickly and briefly; they may be overstated or lack context, and sometimes contain exaggerations or sensationalized claims with the intention of attracting readers to an otherwise reliable article. They are often written by copy editors instead of the researchers and journalists who wrote the articles.
A number of sources are deprecated on Wikipedia. That means they should not be used, unless there is a specific consensus to do so. Deprecation happens through a
reliable sources noticeboard. It is reserved for sources that have a substantial history of fabrication or other serious factual accuracy issues (e.g. promoting unfounded conspiracy theories
), usually when there are large numbers of references to the source giving rise to concerns about the integrity of information in the encyclopedia.
A deprecated source should not be used to support factual claims. While there are exceptions for
fringe view
.
Some sources are blacklisted, and can not be used at all. Blacklisting is generally reserved for sources which are added abusively, such as state-sponsored fake news sites with a history of addition by
Wikipedia:Blacklist
for other details about blacklisting.
See also
Templates
Wikipedia:Template messages/Cleanup/Verifiability and sources
^Many submissions to these predatory journals will be by scholars that a) cannot get their theories published in legitimate journals, b) were looking to quickly publish something to boost their academic resumes, or c) were honestly looking for a legitimate peer-review process to validate new ideas, but were denied the feedback by fraudulent publishers.
^A variety of these incidents have been documented by Private Eye and others and discussed on Wikipedia, where incorrect details from articles added as vandalism or otherwise have appeared in newspapers.
^Please keep in mind that any exceptional claim would require exceptional sources, and this is policy.
^In one case, a lawyer used ChatGPT to generate and file a legal brief that he did not check; the judge upon reviewing the case stated, "six of the submitted cases appear to be bogus judicial decisions with bogus quotes and bogus internal citations", although ChatGPT had assured the author that they were real and could "be found in reputable legal databases such as LexisNexis and Westlaw".[17] Citations have been published to newspaper articles that do not exist, attributed to named reporters.[18]
Salon. Salon Media Group. Archived from the original on October 16, 2011. Retrieved October 17, 2011. Debbie Nathan
also documents a connection between Schreiber and Terry Morris, a 'pioneer' of this [human interest] genre who freely admitted to taking 'considerable license with the facts that are given to me.'
^"Book reviews". Scholarly definition document. Princeton. 2011. Archived from the original on November 5, 2011. Retrieved September 22, 2011.
^"Book reviews". Scholarly definition document. Virginia Polytechnic Institute and State University. 2011. Archived from the original on September 10, 2011. Retrieved September 22, 2011.
^Malone Kircher, Madison (November 15, 2016). "Fake Facebook news sites to avoid". New York Magazine. Archived from the original on November 16, 2016. Retrieved November 15, 2016.
^An example is the Daily Mail, which is broadly considered a questionable and prohibited source, per this RfC.
