Possible "major points and findings" (BROAD focus)

This is a meta-thread, to collate and track Key findings and ideas that appear to have wide traction; Proposals which have a possibility of becoming recommendations (or supporting/secondary recommendations); To discuss their merits as key findings or recommendations; and To debate which ideas should be recommendations.

Themes and ideas from Archive 1:

1. Wikipedia is measured by failures not successes (which are also the easier to fix and get mass participation).
2. Editing Wikipedia shouldn't be a matter of simply signing in and then having total freedom, it should be a learning process that only begins with signing in.
3. Wikipedia has a strong anti-authoritarian component, so suggested changes need to be compelling, pervasive, and persuasive. But equally it has a goal, and we may need to decide some old aspects of ethos are no longer workable in their existing form.
4. Wikipedia needs to revisit the balance of "anyone can edit" and "high quality". This can be made easy and can be nuanced, but ultimately communities need to decide which matters more when they conflict - and back that decision not just "mouth" it.
5. A wizard based interface (with selectable user skill level) is essential to help newcomers and will also facilitate quality.
6. Focus on the "low hanging fruit" - poor quality articles and baseline quality, wide base of newcomers who don't know the ropes, etc. Larger numbers, easier to address, easier to automate, etc. If these are well addressed, higher quality content and established editors will also benefit.

As a gentle prod: have we done sufficient research into non-English wikipedias to be able to say these apply on other language projects? :)

@Philippe: it really depends. As far as I know, all have their own specific demands, problems, guidelines and community mentality. However, in general, the process works the same.

As far as I can judge, the German Wikipedia is the most different from the English one. It has a comparable quality level but the machine seems to produce less "dust". In general, their FA articles (last time I looked, they had more than any other project!) are really well-written, accurate and enjoyable to read. Their total coverage is smaller than that of the English Wikipedia though. It is possible the difference was caused because wp-de is older than any other project except for wp-en. When wp-de was founded, there simply was little to translate/compare with yet at wp-en. Also take in mind that wp-de was not just set up as a German version of Wikipedia, but also as an experiment how to build Wikipedia differently.

Other large projects seem to basically have the same machinery as wp-en but with their own specific niches in content and guidelines. Content niches can be language-, geography- or culture-related (i.e. the Spanish Wikipedia will have better articles about Mexican culture or Spanish cuisine), but also otherwise. Excellent articles on general subjects can exist at (for example) wp-pt or wp-fr, while wp-en only has a stub. The French project has developed its own characteristic lay-out; the Dutch project is focussed on maintenance, etc. In general, I find that (apart from wp-de) most projects translated their guidelines and project pages to some degree from wp-en and have therefore more or less the same community and content evolution, all at their own time and pace. They are dealing or will be dealing with the same problems wp-en deals with (but often less severely because of their smaller size). Some projects may experience problems that have been encountered before by wp-en, but have now been solved there (Yaroslav Blanter gave an interesting overview of wp-ru on this talk page, which shows us it experiences problems that seem to be common for projects with 250,000-1,000,000 articles).

Therefore I assume that FT2's 6 points are valuable for all projects. That being said, we should keep the differences in mind. Differences in guidelines can reflect cultural difference. Is sound perhaps cliché, but my own experience with German people is that they have a perfectionist mentality and are in general more interested in materialistic quality. This seems to be reflected in the content of the German project. I found similar cultural influences when comparing the content and guidelines of wp-fr, wp-es and wp-nl. Weather by chance or because of cultural differences, probably most larger projects did invent or explore new ways of doing things. Whenever one project develops its own way for something, we should analyse what happens and use the knowledge to the benefit of all projects. I'm sorry, but the only project where I found contributors that think they are the only Wikipedia is wp-en. I believe this isolationist attitude can in the end be harmful for that project.

I agree with Woodwalker. We have some expertise in projects different from en.wp, and they seem to have many common features, so that the above points are relevant to some extent for all of them.

On the other hand, smaller wikipedias have markedly different problems. For instance, I have close relations with admins of crh.wp and xal.wp - these are the projects with several hundred articles. They are really happy to accept any non-vandal contribution, and they are very far from the state when all "low-hanging fruit" has been consumed. They have more issues like establishing the proper spelling (on xal.wp the four active native speaker participants had a quarrel about the geographic names of countries, and apparently the reliable sources on this subject do not exist in their language at all - and one contributor left the project since his version was rejected). Obviously our point are not valid for such small projects. To help them, I believe one needs to approach them separately.

I am also ashamed to say I do not have a slightest idea on what happens on the projects other then Wikipedia and whether our discussion apply over there.

I'm glad to see this come up... please continue to research on this. It's important, I think.

Hi everyone. My name is Tyler and I am a member of the Bridgespan team that is helping to support the strategic planning process. I am looking for opportunities to support the great work that this Task Force is doing, and wanted to ask a question related to this thread and the archived materials: Does anyone know of good examples (interal to Wikimedia or external) of other web sites or online projects that are using some of the tools and approaches that are rising to the top of this group's thinking (e.g. wizard based interfaces, automating around low hanging fruit, patrolling, etc). If so, it might be helpful to do a little benchmarking and see if there are any interesting learnings, keys to success, or watch-outs that might be relevant as you continue to push towards recommendations. I would love to contribute to that research if you think that it would be helpful, but need ideas for good places to start!

Hi Tyler, and welcome! :-)

We need to talk to both WikiHow and Wikia about both of these topics, as they have lots of important lessons we can draw from. Let me know if you need help connecting with them. Thanks for all your great work so far!

I think WikiHow attracts a substantially different class of contributors, because their license forbids commercial re-use of editor contributions. I see that as an unfortunate restriction, so I don't contribute there.

I noted this difference in wikiHow.

For the purposes of understanding how the tool can influence quality, I think wikiHow is an excellent model.

Thanks for the suggestions! Are there others we should consider from the perspective of how tools influence quality? What about other collaborative encyclopedias like Hudong?

Here's the information we currently know about Baidu and Hudong. You may want to pose these questions to the China Task force to see if they have insights or can post screenshots.

Tyler - no, I can't think of any (though I wouldn't be surprised if others have designed them).

As the usage here is pretty specialized I'm inclined to say "don't expect to find examples elsewhere". A bit like Mediawiki itself, it's going to be a "home creation" job. I'm happy to do some mockups if that would help.

I see the taskforce as better in coming up with the ideas on functionality/approach/ethos needed, rather than trying to locate specific exemplars. I think we'd be hard pressed on the latter anyway...

And I wasn't logged in. Apologies!

Is there any evidence in support of the proposition that "Wikipedia needs to revisit the balance of 'anyone can edit' and 'high quality'"? If so, what is it?

A pitiful state of a certain class of articles

Wikipedia's been around for 9 years. A large number of users have been around for 4 - 8 years. The impression (and anecdotal impressions are by no means necessarily wrong for the want of a formal research project) relates to users who do patrolling, content issue fixing, tagging, dispute handling, and see the issues close up in detail, and get to know what problems are common; tool writers and admins who get asked for help, etc. The same issues keep coming up. The theme of "anyone can edit" vs "high quality" seems to be quite common in debates on editing process and policy. Asking this head on may be valuable.

But more generally, it's largely anecdotal. In a way, that's the point of this taskforce, to choose people of significant and varied experience and see what they feel and what views and conclusions they reach. Much of what we're doing here is drawing on collective experience because these areas are well experienced but not necessarily well theorized or researched (indeed in complex areas and systems, early research can miss the point that experience quickly highlights -- ie the idea may come first, the research to check it comes after).

That said, if it's not an issue then it won't get much buy-in from the group here as a whole.

Some random ideas :)

What could we try to measure in order to test this notion?

1) The easiest. Take a list of 50 internationally recognized scholars in different field, ask them to pick up two articles in their fields and give a mark 1 to 10 (1 = substandard or non-existent; 10 = excellent quality).

2) More difficult. Tag important specialized articles (may be ask projects to deliver a list) and look at the dynamics of their volume increase. I am sure most of them are small and not expanding.

On WP in en, fr and hu you can look at the quality statistics from the assessment schemes. If you want to locate the poor-quality-but-important articles, you can do this with our new bot that is likely to go live very soon. The new bot will also give an importance score, which uses quite a sophisticated algorithm based on 3 (or often 4) parameters. I would encourage this task force to take a look at what the Elements project on en has done with this sort of information - they often focus their collective effort on specific articles they think are important but most needing work. They use this graphic to show very clearly where the work needs to be done. The result - this small but active WikiProject now has 14 FAs and 18 GAs - about a quarter of all their articles in total, and probably more FAs than all of the rest of chemistry combined (I'm ashamed to say!).

My basic point - if you give a tool to editors that helps them find and track the articles that need work in their area of interest, they will use that tool to focus their efforts. Walkerma 17:55, 6 January 2010 (UTC)