Wikipedia:WikiProject Reforming Wikipedia/Content

Wikipedia has some content of respectable quality to present to its readers: featured and good articles, for example. However, the vast majority of Wikipedia content is not of this quality. Most content is unsourced, and even some core articles have large blocks of unsourced content. Wikipedia admits to its own unreliability, which is hardly something to be proud of. Some of our recognized content degenerates over time, and Wikipedia is still plagued with the problem of readers seeing vandalism, even on the highly visible, showcased TFA. As a result, Wikipedia still has a poor reputation among academics. Wikipedia's completely open new page patrol system allows anyone to patrol articles, even new editors who are rather unfamiliar with our content expectations. This sometimes means that articles of unsatisfactory quality are approved and never attended to again. Quite plainly, Wikipedia has many obstacles to overcome before it is recognized as a serious reference work.

Sourcing and article quality

Issues

The majority of Wikipedia articles are mostly unsourced. Anyone who wants to see this for themselves can simply click through a few random articles, and it will become plain that thoroughly sourced content is, in fact, quite rare in ordinary articles. Although we display rigorously reviewed content on the front page, that is not representative of Wikipedia as a whole—it is these ordinary articles that everyday readers see as a result of their various web searches. Even more disturbingly, some articles deemed to be of vital quality are poorly referenced. A survey of the listed pages will reveal that even articles of Level 1 importance have swathes of unsourced content. Wikipedia is the 7th most popular website in the world, receiving tens of millions of views per day, and its articles almost always rank on top of search engine results. People rely on it every day as the default place to quickly look up information. The widespread presence of unverified material can become downright hazardous, seeing that many people generally accept Wikipedia's information as fact and therefore believe potentially inaccurate information (even on dangerous topics such as medicine). Policy itself admits that Wikipedia is unreliable (also see a relevant essay) and the Reliability of Wikipedia article concedes that "[i]naccurate information may persist in Wikipedia for a long time before it is challenged."

Wikipedia has other quality issues. Some of our recognized content (GAs/FAs) tends to erode and decline over time, as unsourced and poorly written content occasionally slips in unnoticed. Anyone can see featured article review and good article review for examples. It is not uncommon for the Guild of Copy Editors to receive requests relating to the quality degeneration of a GA/FA.

Vandalism is also a very well-known problem. Even if vandalism is reverted in a minute or less, this does no good if the page is an important, popular one that is viewed very frequently. Some reader will arrive at the page in the intervening time and will see the vandalism. We cannot completely rely on our (semi-)automated systems for vandalism prevention—ClueBot does not catch every instance of vandalism, and it sometimes breaks down. Editors speeding through revisions via Huggle are bound to miss things as well. It is a fact that some vandalism stays unreverted for a long time, and there can be consequences (sometimes major) for Wikipedia's reputation (see Wikipedia Seigenthaler biography incident and Wikipedia's very own compilation of incidents). Note that I, the writer of this sentence (Biblioworm) can personally attest to having seen unreverted blatant vandalism on the TFA at least two or three times. It can take surprisingly long to revert vandalism even on the TFA, which is the most viewed page on any given day. It is not uncommon for TFA vandalism to remain for several minutes before being reverted, by which point many people have already seen it. Surely, those readers do not leave Wikipedia with a good impression.

These quality issues are the main reason for Wikipedia's poor reputation among academics—they simply do not take it seriously as a legitimate source for information. It is rather embarrassing that a website which claims to be an encyclopedia (1) is forced to admit is own unreliability and (2) is disregarded among those whose very profession is doing serious research. Clearly, there are basic flaws in the system that must be addressed.

Possible reforms

• Protect recognized content and vital articles with Level 1 PC protection. PC1 protection will still allow everyone to edit the article—it will only require that the edits of IPs and non-autoconfirmed users be reviewed before becoming visible to the general public. Implementing this reform will (1) help prevent the degeneration of recognized content and (2) keep readers from ever seeing vandalism to important and high-quality articles. (If this proves to be successful, we may later consider implementing Flagged Revisions for all articles.) To help our reviewers process revisions more efficiently, we could add an Approve button to semi-automated tools such as Huggle.

• It might also be prudent to improve the requirements for becoming a reviewer. If they will become primarily responsible for preserving article quality, we must be confident that they truly know the content policies. Unfortunately, our current practice is to give this right to pretty much anyone with a few hundred edits.

• Invite experts to review articles, especially in topic areas where amateurs cannot be expected to write a complete and accurate article. Place notices on all articles that experts have peer reviewed, so readers will know that the article is trustworthy. Editor Peter Damian performed a study comparing Wikipedia with the peer reviewed Stanford Encyclopedia of Philosophy. The study clearly showed that the "traditional model wins hands down, at least for the complex and difficult subjects engaged by the SEP." Perhaps specialist expert editors could establish a "stable version" of a Wikipedia article which could be linked to from the current version. We could also consider establishing an editorial board for each subject area.

New pages

Issues

Wikipedia allows any editor with more than 10 edits and four days' tenure to patrol articles. Patrolling is often done by new editors who often pass articles failing our minimum standards, preventing more experienced editors from acting upon them.

Possible reforms

• Inform new users about Wikipedia's bare minimum standard for new articles.

Presentation and accessibility

• Ensure that all content is accessible to at least a wide majority of our readership, especially to people with disabilities such as vision impairment, dyslexia, etc.
• Improve mathematical typesetting, relieving mathematical editors from difficult syntax to produce simple mathematical formulae:

• LaTeX-based typesetting on Wikipedia does not flow well with the text: example: ${\textstyle lim_{\delta {}x\rightarrow {}0}{\frac {(6+\delta {}x)^{2}-6^{2}}{\delta {x}}}=12}$. HTML-based typesetting flows more smoothly; however, it is tedious to typeset formulae using HTML and typographical quality is poor (lim_δx→0 ⁠(6+δx)² + 6²/δx⁠ = 12). A possible solution is to use MathJax; however, developers have refused to implement even optional client-side MathJax in lieu of MathML or inline images: link

Notes and references

References

• Giles, Jim (2005). "Internet encyclopædias go head to head". Nature. 438 (7070): 900–901. doi:10.1038/438900a.
• Read, Brock (2006). "Can Wikipedia Ever Make the Grade?". The Chronicle of Higher Education. 53 (10). A31, pp. 1–9.