Wikimedia Sverige/Lennart thoughts of Quality

Right now we have [roughly 276 000] articles on Swedish Wikipedia.

Hannibal

This essay was originally published in swedish as 10 tankar om kvalitet

Ten possibly provoking thoughts about improving the quality of Swedish Wikipedia

For a long time that has been the most publicized measure of how good we are. Obviously it’s good to have many articles. But already in 2006 Jimmy Wales spoke at Wikimania about how he wanted Wikipedia to go from a quantity based point of view (the maximum number of articles) to a quality based. This you could divide into two phases:

• Phase one, where we’ve been so far, could be likened with ”boldly going where no one has gone before”.
• Phase two, which we are entering now, is rather about seeing to it that the frontier town will get some law and order.

Frank Schulenburg during the meeting in Gothenburg.

During my two days long meeting with Frank Schulenburg from Wikimedia Deutschland in January of 2008, we discussed, in general terms, why the German Wikipedia is often considered one of the best versions of Wikipedia. Last year, in the magazine Stern, there was a comparison between German Wikipedia (dewp) and Brockhaus (the German equivalent of our Nationalencyklopedin), where dewp’s grades were much better than Brockhaus’s. See The front page of the Stern issue where the comparison is published. That could not be said about Swedish Wikipedia—yet. But also not about English Wikipedia either.

One explanation that has floated around is that the English Wikipedia, to some extent, is written by users with mother tongues other than English, which makes the level of the text lower and hence lowers the quality. Since the Swedish Wikipedia in that respect is more like the dewp than the English Wikipedia (enwp), we can leave that theory and instead concentrate on two more important questions:

1. What have dewp done to reach such a high quality?
2. Could the Swedish Wikipedia (svwp) reach the same high quality?

The answer to question number 2 is: Frank thought so. And I think so, too.

The answer to question number 1 is what the rest of this essay is about.

I wish to stress that these are just thoughts, if provoking thoughts. I am aware that all these proposals neither can nor should be implemented. That is not the goal. (I am, for example, not sure that I agree with each and every proposal that I write about.) The goal is instead to get all regular users to think about quality (phase two), rather than quantity (phase one), but also rather than various other considerations. But we will get there.

Quite simply, this is my take on dewp’s recipe to become a better encyclopedia.

Thought number one: delete the bad articles

I want to begin with a controversial proposal. It was controversial on dewp and the result there is still not totally clear, but I think that it may be good to start with a jolt.

The proposal is to remove all bad articles. Bad articles come in many forms: stubs and substubs, articles with low real content (e.g. peacock and weasel terms), articles without proper language, articles that are confusing, lists that can never be completed, etc. By deleting them we won’t have a maintenance page that’s always full of things to do and which has become a constant guilty conscience, rather than a project which sometimes will be more or less fulfilled.

How would this be done? Well, the active users in a particular topic, for example a project or a portal (in other words: people who are interested), regularly go through “their” categories and weed out the worst articles - of course, they should enhance the articles they can”. Enhancements are naturally better, but in many instances it would take such a long time that it’s better to simply remove the articles.

Summary: This deletion proposal would lead to svwp downsizing the article count. We could even go below 250,000 articles. But think of it like this: what a press release we could issue! “Wikipedia takes out all garbage.”

Thought number two: remove all conflicts

They take up a lot of energy, the conflicts, the arbitrations, and the bothersome users that push their agenda and spend most of their time on svwp discussing the topic instead of writing articles.

The proposal, which is also based on discussions with Delphine Ménard, about her experiences with the French Wikipedia, is to remove the conflicts from svwp.

Remove the conflicts? Is that even possible? Most likely there will always be differences of opinion about how articles should read, but lately these conflicts have come to deal more and more with the topics themselves than how the articles should look. It can be fun to take a break from writing articles by discussing something, but it also creates a lot of conflict and draws focus away from the goal of making an encyclopedia. It has become too much “communism is bad” instead of “how should the article about communism be balanced?” (This is *not* solely about how communism should be portrayed, but representative of two different ways of working. One way works on webforums where the goal isn’t anything other than discussing, the other way works better when the goal is to write an encyclopedia.)

What I am talking about is a lower tolerance level:

• Faster blocking for things that are not vandalism (personal attacks, bullshitting, for example). It doesn’t necessarily have to be long blocks, and sometimes warnings are enough - or why not a question on the user’s talk page?
• Letting people know when the discussion goes too far off-topic.
• Telling people to take a wikibreak when they seem to be stressed.
• Focus more on mediation than on getting admins to take action against the other party.

And this will require:

• More admins and less prestige about having the admin tools
• Less tall poppy syndrome (admins, including me, have let irritating behavior through, most likely because of the “should I’s” - I who don’t know the subject, who hasn’t been through every part of the discussion, who recently became an admin, who doesn’t want to make enemies, etc.) and more "can do" (it’s not that hard to know what a personal attack looks like, eh?)
• More barnstars and other forms of kudos for those who take an active part in conflict resolution.

Summary: Think of being able to spend your time here actually making the articles better rather than answering people on talk pages. And every conflict is another risk of alienating yet another active user.

Thought number three: add a quality meter

At the moment, the following message welcomes users at the Main page of svwp:

Välkommen till en:Wikipedia,

den fria encyklopedin som alla kan redigera.

Just nu finns det 151,044 artiklar på svenska

Read it carefully. I don't think this message gives the impression that we focus very much on quality. At all. We just want to show how many articles we have got, and that everybody can edit them.

But maybe that isn't very surprising, since there even isn't a "magic word" to indicate how many featured articles a specific language version of Wikipedia has got. There are magic words to show the number of users, the number of admins, the number of edits, the number of uploaded files, etc, but none for quality. However, we have now created the template sv:Mall:Antal utvalda artiklar, which says how many featured and good articles Swedish Wikipedia has at the moment.

Hence, I'd like to add one thing: a meter that shows the quality of the encyclopedia, already at the Main page. This is an example, that has been created quite quickly (maybe some kind of graph would be nice?), just to show what I mean:

Välkommen till en:Wikipedia,

den fria encyklopedin som alla kan redigera.

Just nu finns det 151,044 artiklar på svenska, varav 1,3‰ är utvalda för sin kvalitet

Then we wouldn't focus just on how many articles we have got, but also on the quality of them, already from the start. By putting the textbox there, we would almost be forced to increase our quality. (And when I say "Main page" I don't mean only the Main page, but for this should be a number that's seen on various places.)

As you can see, we count our featured and good articles in permille, not even in percent. Just imagine what an inspiration it would be to increase this percentage to maybe 1%. (As of today, this would mean that 2700 of our articles were either featured or good. As an comparison, dewp has 1260 featured and 2295 good articles. After all, dewiki is about ten times as big as svwp, but if you see to the percentage of featured articles, we are about equal (dewp's 1,81 permille compared to svwp's 1,3 permille).

The user Danny of the English Wikipedia - the same person that started Veropedia and the competition about best new articles - did write in September 2006, that English Wikipedia should aim for 100 000 featured articles. 100 000 featured articles! Now, that would have been something. (Also see 100 000 feature-quality articles.)

Enwp also has a list of which users that have been the main authors of the greatest number of featured articles, here. Would that be something to introduce just to make more writers to do that last little something? Who is first to reach 10 featured articles? Has someone already reached 10? How about 20 then?

Summary: Just imagine that svwp could reach 1% featured articles. Then we would be five times as good as German Wikipedia!

Thought number four: get money

How Wikimedia Foundation want to spend their money

Wikipedia is nonprofit - I know. But to get Wikipedia work at all, donations are needed. (By the way, please donate here or to bank giro 5822-9915 to support Wikimedia Sverige.) The money goes to servers, bandwidth, technical staff, and the expenditure of the organisation (see planned expenses 2007-2008).

It would be sad though if the donations just would be about getting Wikipedia to work. Donations are also about increasing the quality.

The counterpart of Wikimedia Sverige in Germany, Wikimedia Deutschland, works hard with their money to increase the quality:

• buying expensive reference literature for research of heavy subjects (which later is "stored" by the users)
• travel expenses so that the Wikipedians could meet and discuss projects, or just meet (because it is easier to discuss with people you know - a German Wikipedian who was living in Finland could meet other German Wikipedians for the first time thanks to WMDE)
• prize money in article-writing-contests, for example Zedler-medaljen which is recieved during Wikipedia Academy (dewp has the contest two times a year - 40% of the contributions becomes featured articles)
• arrange fairs and so on, for example Wikipedia Academy
• fix scanning of pictures to Wikisource (other local chapters of WMF sends out photographers to photograph etc.)

More than that: The contributions sent to Wikimedia Deutschland have also led to them gaining more respect - according to the thought "if grant institution X have seen that Wikimedia Deutschland/Wikipedia is good, maybe we should help them as well" (which have led to them gaining more contributions and so on). They have also been able to send out press releases about their successes, which have given them positive PR. It is often just about getting throught the first application procedure, and the next time it is easier to pass.

On the longer run we hope that Wikimedia Sverige will be able to take the same types of initiatives. But we need more ideas of where money should come from, and where they should go. All such proposals/ideas are appreciated. You can also, of course, help. Please sign up on the member list of Wikimedia Sverige.

To dream even more: what if Wikimedia Foundation could recieve enough money (the last fundraising didn't correspond with the expectations and some planned projects are therefore cut down) to spend its energy on more important things than to think about money all the time. Then they would be able to design an interface that is more like the result than today's complicated tags, templates and and tables.

Summary: With more money it would be much easier to get our press releases publicated in media which would increase the interest of editing on Wikipedia. "Wikipedia awards the best article", "She got her student books from Wikipedia", or preferably, "Swede makes Wikipedia simple to edit".

Thought number five: more clearly defined projects

Frank told me about two or three users on dewp who decided to create articles about all lilies (project). They made it. It took about a year. We've had similar projects on svwp: Project Tintin for example, where the goal to create articles was accomplished in less than a few months. After that, some think, the project has reached a stand still. I would say that the project largely has reached its goal.

But there's no shortage of projects on svwp. Presently there a couple of hundred projects. Far from all of them have any clear goal, nor any attainable goal. Not many of them document their progress, something which is good for both the group behind the project and other interested parties. This leads me to a couple of questions: How many of our projects have served their purposes? How many of them are dead? Anyone willing to go to town with the template {{:sv:Mall:Färdigt projekt|This project is done}}?

Recently there was a project on svwp with amazing results, see Project wikify. One thing that I believe contributed to the success of that project is that they calculated what needed to be done and what was done and presented that on the project page. The goal became very clear.

Summary With clearer goals for the projects we can accomplish great things. Then we should dismantle that project.

Thought number six: rally more Wikipedians

Right now there are [57 000] registered users on svwp. We need to be many more if we are to raise the quality in any mentionable measure. As it is now, around half a percent makes roughly 50 percent of all edits. That means that there are 300 very active Wikipedians against 56 500 rather indifferent Wikipedians.

There are several dangers here: intellectual inbreed, reduced article growth, conflicts that can risk the future of the project, wikistress, POV and blindness to the systematic bias, and last but not least, that outsiders begin to think of the users here as a group that's very hard to get into.

Considering that Sweden, Finland and almost all the other countries where Swedish is a large language have such a large portion of the population connected to the internet, we should, with no big problems, be able to rally more users.

Wikipedians meet to discuss Wikipedia Academy: Hannibal, Frank Schulenburg, Moralist (on his knee), Grillo, Boivie, Mnemo and LA2.

Some proposals for rainy or sunny days:

• make the Edit Wikipedia week a bigger event
• mention for your friends and family that you edit Wikipedia (and why) and encourage them to do the same
• ask someone you know to be an expert on a particular topic to read the article on Wikipedia and help make it better
• ask your school if they can't do a project creating articles with the framework of a certain subject, see Using Wikipedia in school
• mention Wikipedia (and the other Wikimedia Foundation projects) as a source when you post on internet forums
• use images and other media from Wikimedia Commons, quotes from Wikiquote, word explanations from Wiktionary, etc, in essays, articles and other written texts
• greet and guide new users in a friendly way - even the not-so-nice ones
• attent a Wikipedia meetup
• attend the local chapter general assembly (and join, of course, a 100 kronors in the case of Wikimedia Sverige is not a lot of money)
• "accidentally forget" the web browser at the main page of Wikipedia when you finish surfing a public computer - and if possible, bookmark it among the favourites
• buy Wikipedia things from Wikimedia Sverige's webshop or CafePress and wear them in a public place
• start a contest around who can produce the highest number of new users.

Any cooperations would be good ideas: what if Amazon.com, national record archives, radio stations, book review sites, MySpace and similar sites linked to the appropriate page on Wikipedia. Then the web traffic (and probably also the number of users) increase dramatically.

Summary: More users equals higher quality. How about "10 000 registered users on Swedish Wikipedia"? Everyone with more than 10 edits.

Thought number seven: educate the general public in how Wikipedia works

Not everybody knows that Commons contains pictures this good. Please let them know.

I have myself started a cooperation with Gothenburg City Library to see if we can do seminars and workshops for the general public. The library have shown great interest. I doubt that other libraries would be less interested: right now there is a great discussion about how Library 2.0 should look and feel and many already work with databases. And in that contact I have also gotten proposals to speak in front of two senior citizens' internet groups about how Wikipedia works. There are more such groups. What if there were 200 senior citizens adding information about their era's movie stars, tools that are almost forgotten nowadays or aspects on our history that we younger people cannot possibly be aware of.

Teachers need lesson plans for how to view material on the internet. There many of us Wikipedians can teach a lot - about Wikipedia's quality programs (version handling, adminship, recent changes, blocking, oversight, etc) as well as source critizism. By getting the teachers on our side, we could potentially get approximately 100 000 new users each year (the mean number of children born each year in Sweden).

Based on Wikimedia Deutschland's material, Wikimedia Sverige is in the process of developing leaflets to hand out and presentations that practically anyone with a month's worth of experience from Wikipedia could do, about how Wikipedia works. Please help with this! Contact me for more information. If you don't want to do the presentationen yourself there shouldn't be to hard finding someone else to do it, if you pay for travel expenses.

Summary: Think about the headline "Wikipedia visits school" or why not "Confront Wikipedia at the library"? "The mean age of Wikipedia is now 55 years" may not be a dream, or is it?

Thought number eight: educate the experts in how Wikipedia works

But it's not only more users we need. We also need expert competence in lots of different subjects to make sure the articles not only scratch on the surface. (That's also one of the disadvantages of having so few regular users: we cannot possibly be experts on everything and hence the articles are less deep than if we could stay in our respective areas.)

So, how do we get more experts to contribute to Wikipedia, other than the proposals I've already mentioned?

One of the workshops during Wikipedia Academy in Göttingen 2006.

One proposal that we are already in the process of making come true through Wikimedia Sverige is staging a Wikipedia Academy. In Germany, France and South Africa these meetings between Wikipedia and the academic world have become very successful and have whetted an appetite for more in the respective countries and chapters. For dewp Wikipedia Academy meant among other things an increase in media exposure, but also more contributions from scientists. The best example is a en:emeritus in en:agriculture who was so fascinated by Wikipedia that he started writing two or three articles a day. So far he's written about 300 of them - on a scholarly level. Since he has a personal image library Commons now have a treasure trove the price of which can hardly be overestimated.

The first Swedish Wikipedia Academy will take place in en:Lund, in cooperation with the university. There you can both make new contacts and make yourself useful (we need everything from organisers to kitchen staff).

Another proposal is getting experts for money (or not) to participate as judges in article writing contests. Through seeing that way how good Wikipedia's articles actually can get, they may be lured into writing for Wikipedia, at least some time or another. Imagine [some local celebrity expert] editing in their expert subject.

A third proposal is approaching expert organizations and ask them to help with their areas of expertise. The article on en:torture could for example need an hours work from a specialist at en:Amnesty International. And if we only give them an introductory course in how to enter sources etc into the article, I do not believe that questions about no original research and bias should be overwhelmingly large. It's definitely in their interest to look good on Wikipedia - and as long as we explain that the best way to get respect is to be neutral, I believe that we can handle the organizations that cannot manage that trust.

Summary: Image the headlines ”Now you can correct [some famous know-it-all] on Wikipedia” or ”The founder of Wikipedia comes to Sweden ― checking the collaboration with Lund's university”.

Thought nine: concentration on the basic articles

Okay, say nothing in this essay will happen: no deletion of bad articles, no lowering of the tolerance level for conflicts, no quality meter, no money, no clearly defined projects (but already a project has started to check articles for relevancy, so the risk of nothing happening here is nil), no massive increase in Wikipedians, no education of either the general public nor the experts - what do we do then?

There is still plenty we can do, on both large and small scale, to increase quality: one of my favourite examples is to make it easier for newbies to edit Wikipedia. That's one area with lots to do. Image yourself to be new to Wikipedia and clicking the edit button. To put it mildly: it's not entirely clear what everything in the edit box here means.

It's not easy to learn what rules apply on Wikipedia. I have for a long time planned to make some big changes to the Community Portal and several other pages listed on the menu to the left of every page. They are a mess.

But that's not what I suggest we concentrate for the near future.

This is how Wikipedia should look. Or?

My suggestion is rather that we take a good look at what people a) most likely need and b) really seem to want. There two tools are the starting point for a bigger project:

1. our quality assessing tables which uses the 1000 articles long list of articles all Wikipedias should have but also includes assessements of how svwp's versions are.
2. lists of our most popular articles

Using these two tools we have identified the most important articles. It's basically these articles that Wikipedia is judged upon.

The project is quite simply to during the near future making sure that a) all 1000 articles in the quality assessing table either attain featured or good quality status, and b) the 200 most popular articles every month at least is presentable (does not have template warning of low quality), but ideally also attain featured or good quality status.

I am aware of the fact that this will be no picknick. But it is important, and if we work together it can happen pretty fast. If everyone of the regular Wikipedians (all perhaps 300) take upon themselves three or four articles to enhance until the end of December 2008, we have reached the goal. It's actually not harder than that.