Wikipedia talk:Pushing to validation

The review panel

Based on the discussion today, I came away with the opinion that probably we should incorporate people both from the WikiProject and one or two outsiders, ideally acknowledged experts. I think the outsiders may bring fresh ideas to the article, as well as all important credibility. It was also suggested that Wikipedians might agree to have their own credentials verified by Wikipedia, to allow these to be used too.

For a few subjects, it would be a good idea to identify subject experts who are keen on the idea of volunteering for the project. For example, focusing on WikiProject Tropical cyclones as usual, we have a specialist forecaster that works in the Hydrometeorological Prediction Center, and has worked previously forecasting hurricanes. We also were positively referenced by Gary Padgett in one of his Monthly Global Tropical Cyclone Reports, which are used by meteoreologists around the world. That can get the ball rolling for our particular project if we take advantage of the opportunity, and set an example for other projects. On a side note, we can also use the {{externalpeerreview}} template to certify these reviews. Perhaps this should be added to WikiProjects' best practices? Tito xd^(?!?) 05:21, 5 August 2006 (UTC)[reply]

As a small part of a conversation that took place after the presentation, I think a leason can be gleaned from the magazine publishing industy, which has been using lay-persons to verify facts of all sort for many decades. A great deal of their effectiveness comes from having a well-definied list of reliable sources to consult on various topics. If it becomes a priority for editors to define a list of places to go for verification purposes for a topic, then literally anyone can be enlisted to do the actual work of checking. This would remove some of our reliance on expertise, which is a limited resource. Twoblackeyes 6 August 2006 2:41 PM EST

I agree with this. Experts will be invalauble in pointing out such things as missing content, but reasonably competent lay-people should be able to do fact checking and source criticism on their own. In my design of the m:Wikicite article validation process, I envisioned fact-checking being done by multiple users with their votes being aggregated. One thing that occurred to me is that, while article fact-checking is in progress, vote/fact-check data should be kept secret so that checkers do not mimic each other in order to skew a vote. This leads to the agreeable outcome of votes having really only two patterns (assuming no major POV controversies are still raging and there is no colusion among voters)- either the results of each fact-check are highly skewed (when most checkers act in good faith) or they are random (when checkers act in bad faith). The bureaucrat in charge of an article's validation can use this fact to speed the process along. If, for example, 5 users said a citation was valid and 2 said it was invalid, then the citation should be further investigated. If 50 said it was valid and 2 said it was invalid, though, then it is highly likely to check out.

In fact, this leads to a further possibility- the fact-check screen could be made to throw in a couple random, bogus citations (this data could be easily pulled from m:Wikicat) to keep checkers on their toes.

Jleybov 22:13, 10 August 2006 (UTC)[reply]

Stable versions

I think that the default version that readers see should be the editable (wiki) version. I think the community would not take kindly to having the wiki articles removed from the main namespace and tucked away out of view. It may be (in the future) that stable versions are created and become very established - at that point we could debate this issue. Walkerma 04:31, 5 August 2006 (UTC)[reply]

See Talk:Calculus for a very recent example of this. Tito xd^(?!?) 05:22, 5 August 2006 (UTC)[reply]

I disagree and think that the default view should be the last validated version, with the current working/scratch version within easy reach, however- maybe via a new tab at the top of the page with info on the number of edits since the last stable version- e.g. "Working Version (7 New Edits)"

The reason for this is that an article's default view is "prime real-estate" and hence will always be a tempting target for vandals, link spammers, and people with various agendas. By making the working version non-default we remove the incentive for much anti-social behavior, and in fact help strengthen the social fabric of the community by making the editing process more deliberative and consensus-driven. Revert wars, for example, are driven by the fact that it is easier to seize an article's default view- even if only temporarily- than to work out disagreements with the opposing side in an edit war. With the default view removed from such contention, though, such conflicts will be much less heated, and indeed there will be an incentive for cooperation, since both sides will realize that in order for their particular changes to make it to "publication" they will have to reach some sort of understanding.

Jleybov 22:13, 10 August 2006 (UTC)[reply]

Yeah, after many discussions I'm coming round to that view as well. Even Jimbo at Wikimania said something like, "we don't want schoolteachers getting pictures of giant penises coming on screen when they use Wikipedia in the school!" Someone from de put it well, calling these stable versions "unvandalised versions", he said "My parents are starting to use Wikipedia - I don't want them getting put off by vandalism." I wonder if they can put the last stable version (or validated version) up on screen, but if you hit the edit tab you get the latest version with a warning over the difference? Walkerma 23:42, 10 August 2006 (UTC)[reply]