Quality Control requires control by independent inspectors

Quality Control requires control by independent inspectors

Edited by author.
Last edit: 10:01, 24 June 2011

Other users have noted several problems with quality: the mountain of junk drifting into English WP, the re-slanting of articles by entrenched w:WP:OWNers, and the bullying to ban editors by cliques/cabals of hive-mind editors. Simply put: w:Quality Control requires control, to stop the junk, to stop the re-slanting, and to stop the packs/cliques of hive-mind editors who all arrive within 2 days "mysteriously" to vote the same opinion on issues. I have started other threads about Quality Control problems:

However, this topic focuses on solutions to the above problems, with some suggestions which might work to reduce the low-quality articles:

  • Steer new editors into welcome-teams: Have new editors join one of "87" welcome-teams to develop requested articles, rather than creating new articles about favorite pet subjects, pet people, or new words just invented. On enwiki, w:Special:NewPages shows a large number of non-notable bio-pages, destined to be deleted. Instead, invite the new editors who would be willing to help and learn, first, before insisting on creating a new article about some rare topic.
  • Force per-article edit-timeouts: I am seriously suggesting automatic read-only, 3-month lockouts for editors who play "keep-away" by controlling and reverting all changes to their "favourite" articles. The rationale has been used for years: if new changes were reallly bad for an article, then other editors would come to rescue and revert the article, without the need for a pack of w:WP:OWNers to jealously guard and watchdog an article to keep all other editors away. Perhaps allow 6 continuous months of control (or 9 months, but not 22 solid months), then those editors would see "view source" (formerly "edit" tab) for the next 3-month period, allowing other editors to alter the article into other directions. In some cases, a w:see-saw effect might emerge, when an article is reverted after 3-month updates in each direction. For efficiency, not all articles (nor all editors) would be restricted, but a user could "nominate" an article, and then nominate some editors to restrict, after their 6-month domination had been reached, allowing a small group of people to offset control by a large group, but not restrict all new editors, some of whom might just oppose the fresh new direction of the article, effectively stopping the changes, if bad enough, as the large group would have done.
  • Punish user offenses by demerits not votes: Rather than allow a pack (or clique) of special-interest editors to "!vote" an editor into a topic-ban or edit-block, an editor could be subject to accumulating demerit points, such as 100 demerits for a w:WP:NPA personal-attack insult, but no harsher than that. An apology could partially offset the demerits, with perhaps 50 merits subtracted. Stop the current practice, where an editor can post one negative remark, and then be shut-out with a 1-month edit-block, issued by an admin with a bad attitude. Instead, an editor would need to repeat offenses, to collect enough total demerit points to reach a level of edit-block, due to having excessive demerits.

In general, institute various forms of control, such as welcome-teams to guide new users into helping develop requested articles, rather than filling Wikipedia with non-notable junk pages. Also, use 3-month edit-timeouts to limit the power of groups who play "keep-away" with articles where other people have been denied edits. Plus, control the punishments allowed by packs of like-minded people, who want to ban or ostracize an editor who disagrees with the pack's hive-mind bias of issues. Such controls could be monitored by quality inspectors, who would note the impact for improving the quality of articles.

Wikid7705:38, 10 June 2011

Who inspects the inspectors?

Tgeorgescu15:29, 12 June 2011

Each individual inspector could be issued demerits, by other inspectors, for improper conduct. It would take some months to determine what the types of conduct problems would be, and how many demerits, such as 100 points for a 3RR sequence of 3 reverts to other changes in a 24-hour period. Admins would have less power, on a daily basis, as merely issuing demerits or adding the opposite merits points, and so blocking of users would be relatively rare. Also, to reduce gangs, inspectors could volunteer into inspector pools, as inspector juries to be assigned, at random, within various topic areas they like. Only some inspectors, from a "physics pool" would be assigned to judge the quality of article "w:Isaac Newton" during the current quality period. I think inspectors should get mandatory 3-month time-outs from articles, just as the editors would be denied talk-page edits to an article for 3 months, once their 6-month? domination limit had expired. Private discussions could be encouraged, using outside email accounts, to debate issues without people being "shown wrong" in talk-pages viewed by thousands. A lot more chit-chat hashing of issues could be promotedly, without clogging the servers, if outside email/IM were used more. Being chosen from random pools, there would be less chance that inspectors were the current POV-pushing w:WP:OWNers of the articles.

Meanwhile, I think the idea of new editors joining a welcome-team would scare away many would-be troublemakers, so the current blocking of corporate-gogetters would be reduced, as they would quickly leave once they knew their company products were off-limit articles. Also, outside inspectors could monitor how many users were getting demerits within certain welcome-teams, and perhaps the "top-ten most demeriting" teams could be checked for overly-hostile treatment of other editors. By "quantifying" disagreements as edit-war demerits, or name-slur demerits, rather than capricious blocks, at the whim of irritated admins, then patterns of conflicts could be better spotted as patterns in the numbers assigned to editors (and articles) in each team. Once the quality of articles is quantified as specific quality metrics (number of words, number of grammar errors, number of prepositions, number of footnotes, etc.), and the quality of user actions is quantified by adverse demerits or rewarded merits, then the system would begin to reveal obvious patterns in the numbers being measured about those qualities.

BOTTOMLINE: Quality control requires control, but such control should be randomized from pools of interested users, where editors could no longer just gang-up on a particular article because they have buddies to force their POV onto the article. Hence, an organized system of measuring article qualities and user-action qualities would be instituted to replace the current winners-take-all games which play "keep-away" with articles, where new editors are denied from making the changes.

Wikid7710:43, 24 June 2011

The idea is that admins do such sort of jobs, and inspectors would be placed above the admins. Now, there are not many admins, and the things would get excessively complicated in order to pamper the users who cannot adapt to the Wikipedia editing climate. You ignore how much work have to do the people who check the recent changes. They do not seek to bully others, but simply keep the articles free from spam, rants, vandalism, fringe theories, racism, insults, libel and so on. Wikipedia articles would degrade considerably in several days if these people would not do their job. Wikipedia has clear policies about what can be added or removed from the articles and has arbitration committees and mediation procedures. These should be enough to any person who is prepared to abide by the rules.

Tgeorgescu20:09, 25 July 2011
 
 
Empty the bullies of Wikipedia!
Vidons les intimidateurs de Wikipedia!
Jeff17:36, 17 July 2011