Wikipedia:Assessing reliability

Because anyone can edit Wikipedia articles, critics have questioned the extent to which information in Wikipedia articles should be considered trustworthy and actually factual. This essay has been created to help readers assess the reliability of articles.

It should be noted that this essay cannot be used to prove any particular article should be considered highly reliable; the General disclaimer still applies.

The table below is an overview of reliability indicators. More information on each one can be found after the table.

Verifiability
Verifiability is probably the most accurate way to assess the reliability of an article. The following are indicators of verifiability.

Number of Sources
At the bottom of a good article, a section, usually called "References" or "Notes", will list sources that were used in writing the article. If this list is extensive the article is generally reliable.

Articles of high reliability will often contain both online sources (freely accessible via the Web) and offline sources (books or scholarly articles). The presence of online sources makes it easier for users to verify factual accuracy, whereas the presence of offline sources may be an indicator of thorough research.

Some examples of good reference lists include the articles Albert Einstein and Facebook.

Reliability of sources
Particularly if there are few sources listed, it is a good idea to scan those sources to see if they are reliable. Unreliable sources include most blogs and personal websites, although some of these are reliable. Reliable sources include most newspaper and magazine articles, online encyclopedias, and other professional publications such as books. Sources do not need to be online to be cited. See Evaluating sources for more information on determining reliable sources.

Inline citations
In order to verify specific facts and controversial statements, inline citations are used to match sources in the reference list with facts in the article (Example 1).

On the other hand, inline text is also used to mark facts where verifiability is needed (Example 2) and where it is suspected that a statement is original research and is not verified by any source (Example 3). When you see these, it is a warning that the specific fact is unverifiable. On the other hand, it means that contributors of the article are keeping verifiability in mind.

Examples
Example 1: This is a fact.

Example 2: This appears to be an unverified fact that needs a citation.

Example 3: This might be original research..

Verifiability tags
The most obvious sign of verifiability problems is when an article is tagged for verifiability issues.

Examples
The following are some verifiability tags (templates) used on articles. More can be found here.

Stability
Stability is important because if an article is stable, it usually means the article is watched by many editors and therefore poor information will not go unnoticed. There tends to be a correlation between the stability of an article and its age, its length, the number of different contributors to the article, and the type of editors of the article.

Age of article
The older an article is, the more likely it has been scrutinized for verifiability, possible deletion and other factors. For example, an article that is a day old probably will have less reliable information than an article that is 2 years old. An article can be checked for its age by clicking the history tab at the top of the page, clicking "Earliest", and pressing to see the date of the first edit.

Length of article
The lengthier an article, the more likely that there have been good contributions by many editors. However, this does not indicate high reliability if the article is young, or if there have been only a few editors.

Number of different editors
If there are many editors, this means the article has come under more scrutiny. The number of editors can be found by looking at the page history (accessed by clicking the "history" tab at the top of the page). If you see many editor names, then the article can be considered more reliable than an article with few editors. Remember that you're only initially looking at 50 edits; you can click "500" to see a longer list of edits.

Type of editors
If an article's editors are inexperienced they are more likely to be unfamiliar with Wikipedia's policies and guidelines. While they may have provided accurate information, their relatively few postings to Wikipedia may mean that they aren't committed to following Wikipedia's rules; they certainly have no reputation to protect.

Also, contributions by IP addresses (a series of numbers instead of a name, sometimes referred to as "anonymous IP editors"), are more likely to be incorrect (about a quarter of such edits are immediately reversed by other editors for one reason or another); among other reasons, there is even less accountability if someone edits without logging in. Good information can be provided by anonymous IP editors, but any information should be considered less reliable unless a source is provided.

Unfortunately, there is no easy way to see what any given information in an article was added by what editor. Nor is it easy to assess the experience of registered editors. An experimental tool called the Wikipedia trust colouring demo has been introduced to see the reliability of certain areas in an article. Unfortunately, the pages it has rated are not current.

Nature of subject
Sometimes, the nature of the subject of the article will make it inherently less reliable than other articles. Articles are generally less reliable when the subject is controversial or prone to vandalism.

Controversiality
Controversial subjects are prone to edit wars, bias, and generally rash edits. As a result, the information in controversial subjects may often be less reliable.

For example, articles such as Bible, Intelligent Design and An Inconvenient Truth are controversial topics which may experience issues.

However controversial subjects may be highly reliable because they may be scrutinized by valued editors simply because of their nature.

Vandalism prone
The following types of articles may be vandalism prone and therefore less reliable:
 * Commonly visited articles
 * Articles of an explicit nature
 * Articles likely to be edited solely for vandalism, such as Gay and Penis
 * Articles about common subjects learned in school, such as George Washington

Systematic Bias
The average Wikipedian on English Wikipedia is (1) a man, (2) technically inclined, (3) formally educated, (4) an English speaker (native or non-native), (5) white, (6) aged 15–49, (7) from a nominally Christian country, (8) from an industrialized nation, (9) from the Northern Hemisphere, and (10) likely employed as an intellectual rather than as a labourer (cf. User survey and University of Würzburg survey, 2005). As a result, it is natural for systematic bias to form.

Certain articles will be more susceptible to systematic bias. For example, the article Uses of torture in recent times tends to display information about few, but well documented, Western World cases. It neglects to document torture cases in many other countries around the world.