Wikipedia:Potentially unreliable sources

Wikipedia's requirement for writing articles is "verifiability, not truth". We rely on what is written in external sources to write this encyclopedia, yet not all sources are equal. The guideline Identifying reliable sources gives general advice on what is and isn't a reliable source; this essay aims to analyse specific examples of sources that might initially appear to be reliable, yet may not be. If in doubt about a source, discuss this at the reliable sources noticeboard.

News media
All mainstream news media can make mistakes. Particularly with breaking news, corrections will need to be made and should be watched out for, and much tabloid journalism will be sensationalist and gossip-driven. Fact checking has reduced generally in the news media over recent years. For more on the trend of churnalism, see Flat Earth News, a book by Nick Davies. Specific examples to treat carefully include:
 * State-associated or state-controlled news organisations, especially state media in countries with low press freedom, such as the Chinese press agency Xinhua, the North Korean Korean Central News Agency and Press TV in Iran. They may be propaganda organisations. RT, formerly known as Russia Today, and other Russian government-funded sources like Sputnik News have also been described as propaganda outlets for the government. However, such sources may be reliable for determining the official positions of their sponsoring governments. Similarly, Voice of America, Radio Free Europe/Radio Liberty and other US state media sources may also be unreliable as to facts, as they have been described as propaganda, but may be reliable regarding the official position of the United States.
 * TMZ has received criticism for errors in breaking news and has a reputation for gossip, but it is increasingly seen as credible by other news agencies (1, 2, 3).
 * The more extreme tabloids such as the National Enquirer should never be used, as most stories in them are intentional hoaxes.
 * In general, tabloid newspapers, such as The Sun, Daily Mirror, the Daily Mail (see also the February 2017 RFC discussing its validity), equivalent television shows, should be used with caution, especially if they are making sensational claims. The Daily Express and Sunday Express should be treated with even greater caution.
 * Forbes.com – although a branch of the Forbes magazine, most of its articles are written by paid "contributors"—similar to a content farm (see below). Articles written by Forbes.com contributors are not reviewed by editors prior to publication. A small proportion of Forbes.com contributors may have credentials to qualify for citation under the self-published source criteria (established expert on the subject matter, whose work in the relevant field has previously been published by reliable third-party publications, but must be used as third-party citations on statements relating to living persons).
 * The articles on PR Newswire and VerticalNews are entirely corporate press releases. Other online sources of news rely on such sources for their own articles—even CNN Money and Yahoo! Finance have published some very implausible corporate press release material without telling readers they used text from a site which publishes companies' press releases almost word for word. Any article citing PR Newswire, VerticalNews or similar online business news sources should be considered a primary source unless there is evidence—not in the byline but the body of the article—of independent authorship and editorial review in the article you're citing. Searching the subject of the article you're citing may turn up identically or almost identically-worded articles elsewhere and at that point it's evident that the text of a primary source is being reproduced in the article(s) you're looking at. See WP:RS for when it's good to use primary sources and independent secondary sources.
 * Never use conspiracist sites such as Infowars without exception.
 * Breitbart.com, which has a long and documented history of publishing misrepresentations, fabrications, half-truths and outright lies about people it politically opposes. See the site's article for examples and see the September 2018 RFC that deprecated its use as a reference for facts. May be useful for discussing opinions, but should never be used to support negative claims about people.

Science churnalism sites
Churnalism is the practice of lightly repackaging press releases and republishing them. These sources are WP:SPS and not independent:
 * Phys.org
 * Eurekalert
 * ScienceDaily

Stock chasing blogs
These are blogs that are opinion-driven and subject to all kinds of external interests and speculation. Not what we should be reaching for, with our mission to provide the public with articles summarizing accepted knowledge.
 * Seeking Alpha
 * TheStreet.com
 * The Motley Fool

Sites that may appear to be reliable sources for Wikipedia, but are not

 * Content farms – these include sites such as Examiner.com (not to be confused with the San Francisco Examiner) and those owned by Demand Media. While they may resemble the format used by legitimate websites (especially in the case of the former), the content is by amateur writers paid by page views and other factors, and are effectively self-published, user-generated content that lacks editorial oversight. (see 1, 2, 3)
 * Articlesnatch.com
 * The Onion – In a few high-profile incidents, major news services have reported on content from this satirical news site, mistaking it for real news.
 * The Daily Currant – Satirical news originating on this site mistakenly ended up on a few US news sites.
 * The Lapine – a satirical news site in Canada
 * Newslo.com and Politicalo.com – satirical articles based on actual events that provide a button readers can use to highlight the portions of an article that are real
 * American College of Pediatricians – publishes from an unscientific viewpoint
 * Other sites on the List of satirical news websites

Funeral homes
Obituaries published by funeral homes are the same as an advertisement; the only difference from a commercial advertisement in a glossy magazine being that instead of a corporate sponsor, the ad is being published by the family or friends of the deceased. Examples:
 * Thomas Funerals

Scholarly journals
Scholarly journals are normally reliable sources, but some journals have a reputation for bias or unreliability. QuackWatch has a list of non-recommended periodicals, however, a short list of journals which should be used with extreme caution include:
 * Journal of American Physicians and Surgeons (JPandS), publishes from an unscientific viewpoint
 * Rivista di Biologia, edited by Giuseppe Sermonti, noted for publishing fringe theories
 * Medical Hypotheses, non-peer reviewed and known for unscientific content
 * Energy & Environment, edited by and published by climate change denialists
 * Medical Veritas: The Journal of Medical Truth, published by Medical Veritas International Inc., listed by Quackwatch as a "questionable organization".
 * Mankind Quarterly
 * Society for Scientific Exploration, and its journal, the Journal for Scientific Exploration, which publishes almost exclusively on fringe topics, promoting pseudoscience and conspiracy theories.
 * Any publication with a fringe topic in its name should be treated with caution: most only serve to promote that topic and are not reliable sources for anything other than their own viewpoint. Examples of such promotional journals include Creation Research Society Quarterly, Evidence-based Complementary and Alternative Medicine, and Homeopathy.
 * Be aware of predatory publishers, for example journals published by OMICS Publishing Group. These are very unlikely to be accepted as reliable sources.
 * Be aware of journals which boast about their high impact factor, as they typically reach such status by screening submissions for novelty and other features which are also risk factors: at least in the hard sciences, the articles they publish were shown to struggle to reach average reliability.

Wikipedia mirrors
Wikipedia should not cite itself, but circular referencing and fact-laundering are possibilities if we are unaware that sources we use copy from Wikipedia. Lists are at Republishers and WP:MIRRORS. Some examples that appear in Google Books and are frequently inadvertently used by editors are:
 * Alphascript Publishing and the many other imprints of VDM Publishing (see WP:ALPHASCRIPT, search for uses)
 * Webster's Quotations, Facts and Phrases by Icon Group Publishing (search for uses Mirrors and forks/Ghi)
 * Books LLC (Mirrors and forks/Abc)
 * Multiple Indian books such as Encyclopaedia of India, Pakistan and Bangladesh, Freedom fighters of India, a series of "Faith & philosophy in..." books and some other books by Om Gupta published by ISHA books and Gyan publishing house. See Reliable_sources/Noticeboard/Archive_37
 * Books published by Gyan Publishing / Isha Books are not WP:RS in general. Several of their books dating to a period before Wikipedia have also been plagiarized from other sources. See discussions at WP:RSN, WT:MIRROR, WT:IN etc.
 * (Mirrors and forks/Def)
 * Cram101, aka CTI Reviews, Content Technologies Inc., Just The Facts 101, Textbook Key Facts search for uses
 * Hephaestus Books, search for uses

You can use this note to let editors who added these sources know why they should not be used, and you can use Wikiblame to find when the source was first added.

Online mirrors

 * New World Encyclopedia (search for uses) — an online encyclopedia that, in part, selects and rewrites certain Wikipedia articles through a focus on the values of the Unification Church of Sun Myung Moon. It "aims to organize and present human knowledge in ways consistent with our natural purposes."
 * enotes.com/topic, worldlingo.com, absoluteastronomy.com, spiritus-temporis.com, Conservapedia, revolvy.com,
 * World Heritage Encyclopedia (worldheritage.org), also hosted on World eBook Library (www.ebooklibrary.org) and World Library (www.worldlibrary.org); search for uses.

Online sources

 * h2g2. Wikipedians often make the mistake of thinking that because this used to be hosted by the BBC, it is reliable. It is user generated, and not reliable as a source, though in certain contexts it might meet the criteria for an external link (search for uses).
 * BBC Music. The artist biographies are usually taken directly from Wikipedia, which is clearly indicated on the page.
 * fantasticfiction.co.uk. Used on 1000s of articles about books, but it is a commercial site with no clear editorial oversight. See the Administrators' Noticeboard discussion.
 * Answers.com
 * Wisegeek.com. WP:RSN discussion has described it as a "content farm" that pays its writers to produce "breezy, popular interest pieces with no footnotes" based on popular search terms, and concern was expressed that it may be drawing uncredited information from Wikipedia and creating an information loop.
 * groups.google.com (and other Usenet portals). The quality of Usenet varies, with a large proportion of it being user generated content, with little editorial control (moderated groups being the exception). Usenet threads from such portals may also have been edited compared to original postings. Usenet postings from such portals should not be used for Wikipedia purposes without additional sourcing from reliable non-Usenet sources.

Self-published books
These may appear to be reliable as they are in Google Books and Amazon, but they have no editorial oversight. Some of the biggest self-publishing houses are:
 * iUniverse (search for uses)
 * Lulu.com (search for uses)

Who's who scams
A Who's Who scam is a fraudulent Who's Who biographical directory. While there are many legitimate Who's Who directories, the scams involve the selling of "memberships" in fraudulent directories that are created online or through instant publishing services. Because the purpose of the fraud is only to get money from those included, the contents are unlikely to be reliable.

Fansites
Fansites are generally not considered reliable. However, exceptions can apply - some fan sites contain scans of small extracts of old newspaper and magazine articles, and these may be the most convenient way to cite facts based off the original published content. Be careful, however, as these scans may actually be a copyright violation, which must not be used to cite facts in an article. If using a copyrighted source from a fan site, the citation should be to the original copyrighted source, not the fansite, and the fansite should not be linked to from Wikipedia, not even as a WP:Convenience link. However, be aware of WP:Citing sources - unless the complete source is available, excerpts may be taken out of context, or changed to fit the site's POV, and are therefore unreliable. Transcripts of content are generally not reliable unless produced by a reliable source.

The opinions of a fan site owner or owners are generally not reliable - anyone can set up a web site and claim to be part of an "editorial team" without establishing a widely known reputation for fact checking and content control.

Personal communication
It is a convention in scholarly works to add notes of "personal communication" or "pers. comm." with an individual or organisation who are considered knowledgeable on a topic, e.g. see [http://www.ncbi.nlm.nih.gov/books/NBK7242/ Citing Medicine: The NLM Style Guide for Authors, Editors, and Publishers. Chapter 13: Letters and Other Personal Communication]. On Wikipedia this is considered to be original research, which is not permitted.
 * Search for uses here or here.