Wikipedia:Reliable sources/Perennial sources/Further classification
 | This is an information page. It is not an encyclopedic article, nor one of Wikipedia's policies or guidelines; rather, its purpose is to explain certain aspects of Wikipedia:Reliable sources/Perennial sources. It may reflect differing levels of consensus and vetting. |
On the page WP:Reliable sources/Perennial sources, a common classification is "No consensus, unclear, or additional considerations apply" (also sometimes called WP:MREL, "Option 2", or "yellow" sources). These sources are grouped together because they are not generally reliable or generally unreliable, and because an editor looking to use such a source should take the same general approach: considering it more carefully beforehand, with reference to the summary sections at WP:RSP#Sources. However, within this category, there are still substantial differences that can impact source evaluation. This page provides a taxonomy of MREL sources along with explanations and further information.
The content of this page is based on the summaries of each source at WP:RSP#Sources, where the full descriptions can be found. In the event of a conflict between this page and the main RSP page, defer to RSP.
Types of MREL sources
[edit]A source may be classified as MREL for many reasons, but these reasons generally fall into one or more of the following groups. It is common for individual sources to qualify for multiple groups, e.g. a source that is unreliable on some topics but marginally reliable on others would fall in both Group 1 and Group 2.
Group 1: There is variation in reliability within the source
[edit]The reliability of these sources differs when citing different parts of the same source, or when the source is used to support different types of statements. While this occurs in all sources, this group focuses on cases that are specific to the source in question (e.g. a newspaper that includes both news and opinion does not qualify on that basis, if there are no additional considerations for the specific source). The scale of the issue is also relevant: if a source's issues are confined to a narrow category (e.g. one contributor out of many is later found to have plagiarized their work), it is likely to be considered generally reliable, while if only a small subset of the content is reliable (e.g. Youtube videos posted by reliable sources are themselves reliable), it is likely to be considered generally unreliable. The source may be divided into multiple entries at RSP.
Guidelines for use:
- Assess the source based on the attributes that are described in the RSP summary as producing different degrees of reliability.
Examples:
- Sources that are reliable in some topic areas but not in others
- Sources that are reliable for some types of statements but not for others (e.g. a source only being reliable for uncontroversial information, or not being reliable for statements on living persons)
- Sources that are reliable during certain time periods but not in others (e.g. due to a change in editors)
- Sources that are partially self-published and partially not (e.g. a website that also publishes content written by the owners)
- Sources that have multiple authors, some of which are reliable and others which are not (e.g. when they are subject-matter experts speaking in their area of expertise)
Group 2: The source is marginally reliable
[edit]These sources have been determined to be less than generally reliable but more than generally unreliable (sometimes summarized as "usable with caution"). There are substantial differences in reliability within this group; some of them are close to generally reliable, and others are close to generally unreliable. The evaluation of marginal reliability does not necessarily apply to all aspects of the source (if the source is also classified as Group 1). If the source is not classified as Group 1, other factors may still increase or decrease its reliability in the usual manner, given the specific context.
Guidelines for use:
- Determine the reliability of the source for the specific context. Each use should be evaluated individually, with reference to the RSP summary.
- If included, the source often requires in-text attribution. Some require it all the time, while others may or may not depending on context.
- It is often a good idea to search for a better source.
Examples:
- Sources that are (or are sometimes) biased or opinionated, but do not generally report direct falsehoods
- Sources that often lack the due weight to be included in articles
- Sources that are given some extra latitude when they cover topics that aren't covered by more reliable sources (e.g. see the summaries for China Daily and Daily NK, though this justification has been cited in other discussions as well)
- Sources that have some editorial oversight, but the quality is low or uncertain
Group 3: The status of the source is unclear
[edit]For these sources, the proper evaluation of reliability under Wikipedia's policies and guidelines is uncertain. This is almost always because editors found no consensus when they evaluated the source. As with Group 2, this evaluation does not necessarily apply to all aspects of the source.
Guidelines for use:
- Follow the same guidelines as Group 2.
Examples:
- Editors who have evaluated the source found no consensus
- Editors who have evaluated the source found that a proper evaluation of reliability is impossible (e.g. not enough information is available to make a judgement)
Group 4: The source is used to report opinions
[edit]This includes sources where the reporting of opinions is a primary use case. Reliability may vary depending on the specific author (and the source would also qualify for Group 1), or a source with lower reliability may be acceptable under the lower standards for opinion content (and the source would also qualify for Group 2). This group does not include sources solely because the RSP summary describes them as biased or opinionated, or mentions that the source contains opinion content.
Guidelines for use:
- Determine reliability and weight based on the author (whether a person or an organization) and the editorial process that was applied.
- Apply the usual procedures for opinion pieces, newsblogs, and subject-matter experts where appropriate.
Examples:
- Sources that primarily consist of opinion content (e.g. Jacobin, The Spectator)
- Sources that are sometimes used to report their own opinion (e.g. Cato Institute, other think tanks)
Group 5: The source is only reliable for specific primary-sourced information
[edit]This includes sources that are useful as primary sources but are not reliable as a secondary source. The definition of this group may evolve as more examples are identified; the current members are limited to specific use cases but are reliable within those limits. A more broadly applicable primary source would likely count as generally reliable, while a primary source usable on rare occasions would likely count as generally unreliable.
Guidelines for use:
- Follow the guidance on primary sources while remaining within the relevant use case.
Examples:
- There are currently only two sources in this category: Hansard and Social Blade.
Group 6: The source is no longer updated
[edit]This includes sources where a lack of updates may be relevant to reliability, e.g. a source may no longer be able to issue corrections. As with the other groups, the focus is on sources that may not be fully addressed by the usual procedures, as indicated by inclusion in their RSP summaries. (For example, most books are not updated, but this is already widely understood.)
Guidelines for use:
- Evaluate whether the information being cited could be out of date; if so, consider including the publication date in the text (e.g. using "As of..."), or search for for a more recent source.
Examples:
- Sources that were regularly updated, but no longer exist or no longer publish new content (e.g. a newspaper that goes out of business)
- Sources that are updated inconsistently
Other sources
[edit]Since this classification system is descriptive, it is possible that additional categories will be added in the future.
Approximate categorization of individual sources
[edit]This is an approximate summary of existing MREL sources based on the descriptions from RSP. As noted above, in the event of a conflict between this page and RSP, defer to RSP.
Some classifications only apply to specific parts of a source; details can be found in the relevant RSP entries. When a source has more than one RSP entry, only the MREL entry is included here; such cases are classified in Group 1 by default, although they may be in other groups as well.
| Source | Group 1 (Variation) |
Group 2 (Marginal) |
Group 3 (Unclear) |
Group 4 (Opinions) |
Group 5 (Primary) |
Group 6 (Updates) |
|---|---|---|---|---|---|---|
| Academic repositories |  |
|
||||
| Al Arabiya | |
|
||||
| Al-Manar (general topics) |
|
|
|
|||
| Alexa | |
|
| |||
| Allsides | |
|
||||
| The American Conservative | |
|||||
| Anadolu Agency (general topics) |
|
|
||||
| Anti-Defamation League (antisemitism, excluding Israel or Zionism) |
|
|
||||
| Apple Daily | |
|
| |||
| Arab News | |
|
||||
| Asian News International | |
|
||||
| AskMen | |
|||||
| ARDA | |
|
||||
| ASPI | |
|
||||
| Ballotpedia | |
|||||
| Bloomberg profiles | |
|
||||
| Boing Boing | |
|
||||
| Bustle | |
|||||
| Buzzfeed | |
|||||
| Cato Institute | |
|
|
|||
| CEPR | |
|
|
|
||
| China Daily | |
|
||||
| CBN | |
|
||||
| CNET (October 2020-October 2022) |
|
|
||||
| Cosmopolitan | |
|
||||
| The Daily Beast | |
|
|
|||
| Daily Mirror | |
|||||
| Daily NK | |
|||||
| Daily Telegraph (transgender topics) |
|
|
|
|||
| Democracy Now | |
|
||||
| Destructoid | |
|
||||
| Dexerto | |
|||||
| Dotdash Meredith | |
|
||||
| Encyclopedia Britannica | |
|
||||
| Encyclopedia Astronautica | |
|
| |||
| Entrepreneur | |
|
||||
| Evening Standard | |
|||||
| FAIR | |
|
|
|||
| Fox News (news excluding politics and science) |
|
|
||||
| The Free Press | |
|||||
| Genius | |
|
||||
| GEOnet Names Server (names and locations) |
|
|
||||
| Google Maps | |
|||||
| The Green Papers | |
|||||
| The Guardian blogs | |
|
||||
| Guinness World Records | |
|||||
| Hansard | |
|
|
|||
| Heavy.com | |
|||||
| Hope Not Hate | |
|||||
| HuffPost (politics) |
|
|
|
|||
| Human Events | |
|
|
|||
| IJR | |
|
||||
| Insider (excluding culture) |
|
|
||||
| IslamQA | |
|
||||
| Jacobin | |
|
|
|||
| Jewish Chronicle | |
|
|
|||
| Jezebel | |
|
|
|||
| Kotaku (pre-2023) |
|
|
||||
| Mashable (non-sponsored content) |
|
|
||||
| MDPI | |
|||||
| Media Matters | |
|||||
| Mediaite | |
|||||
| MetalSucks | |
|
||||
| MEMRI | |
|
||||
| MEMO | |
|
|
|||
| Mondoweiss | |
|
||||
| The Morning Star | |
|
|
|||
| National Review | |
|
|
|||
| The Needle Drop | |
|
|
|||
| New York Post (entertainment) |
|
|
||||
| Newsweek (2013-present) |
|
|
||||
| The Next Web | |
|
|
|||
| La Patilla | |
|
||||
| Polygon (May 2025-present) |
|
|
||||
| Pride.com | |
|||||
| Quackwatch | |
|
||||
| RFE/RL | |
|
|
|||
| RealClearPolitics | |
|
||||
| RhythmOne | |
|
||||
| RIA Novosti | |
|
|
|||
| Salon | |
|
||||
| Science-Based Medicine | |
|||||
| ScienceBlogs | |
|
||||
| Screen Rant | |
|
||||
| Sherdog | |
|
||||
| The Skeptic’s Dictionary | |
|
||||
| Sky News Australia | |
|
||||
| Social Blade | |
|
||||
| The South African | |
|
||||
| Spark Notes | |
|||||
| The Spectator | |
|
|
|||
| Sports Illustrated (June 2019-present) |
|
|||||
| The Straits Times | |
|
||||
| TechCrunch | |
|
||||
| TED | |
|
||||
| ThinkProgress | |
|
| |||
| The Times of India (post-1950) |
|
|
||||
| TMZ | |
|
|
|||
| Townhall | |
|
||||
| TRT World | |
|
||||
| Us Weekly | |
|||||
| Vice Media | |
|||||
| Washington Examiner | |
|
|
|||
| The Washington Times | |
|
||||
| Wikidata transcluded statements | |
|||||
| World Christian Database | |
|||||
| WSWS | |
|
|
|
||
| XBIZ | |
|
||||
| Xinhua | |
|