Foundation reports editor trends, technology plans and communication changes

The Wikimedia Foundation's "Editor Trends Study" (commissioned in October "to help better understand the internal dynamics of our communities") was published last week. The summary lists five "early conclusions:":

In a letter to Wikimedians ("March 2011 Update", advertised via CentralNotice – the English version has so far received over 18000 views), the Wikimedia Foundation's Executive Director Sue Gardner presented the study's results, interpreting them as follows:

She then went on to name "Openness Begets Participation" as a strategy to solve the problem ("I believe we need to make editing fun again for everybody: both new editors and experienced editors. ... Quality and openness go hand in hand"), and outlined "The Year Ahead", based on the Foundation's "Product Whitepaper" (Signpost coverage, "a comprehensive analysis of our product priorities" based on its 2010–15 strategic plan (with "product" being defined as "technology through which people receive and develop Wikimedia content"). The following priorities were named, some of them comprising already ongoing efforts:

1. Create a visual editor
2. Improve the newbie experience
3. Support community growth in developing countries
4. Serve audiences on all devices
5. Create a delightful experience for contributing and reviewing multimedia

In related news, Kevin Rutherford recently analyzed the edit numbers of the most active editors on the English Wikipedia (by edit count), concluding that 27% of all edits have been done by a core group of 4,000 editors. In the last three years the number of edits needed to get onto List of 4,000 Wikipedians who have done the most edits rose from 5,000 edits to 11,426. Since the start of last year the number of editors breaking the 100,000 barrier has jumped by more than half, from 68 to 109, whilst the number of editors who have contributed over 200,000 edits has doubled from ten to twenty. This prompted a proposal to broaden the list from the 4,000 editors with the highest edit count to 5,000. So as of 9 March 2011, all editors with 9,168 edits or more are on the latest list (though some have opted out of being named).

Last week, the Wikimedia Foundation posted a job opening for a Movement Communications Manager, a new position (reporting to the Head of Communications, currently Jay Walsh), whose purpose will be to "serve the Wikimedia community and Wikimedia Foundation staff by increasing the quantity and quality of communications between and among the Wikimedia Foundation and the Wikimedia community." The list of job duties, apart from those concerning the communication of the achievements of Wikimedia projects to an external audience, also indicates planned changes to the interaction between the Foundation and volunteers:

Another job opening was posted for a Data Analyst and Researcher‎ who "on a day-to-day basis ... will create, mine and analyze data to help understand readers and editors of different Wikimedia projects, especially Wikipedia, across different geographies", and based on them create "reports, charts, graphs, maps and tables" to "concisely, clearly and meaningfully convey information for a lay audience, Wikimedia community and the foundation."

In other staff news, Veronique Kessler, who has been the Foundation's Chief Financial and Operating Officer (CFOO) since February 2008, announced she will leave for family reasons at the end of June, after completing the development of the 2011–12 business plan.

• GLAM trip to India: Liam Wyatt (User:Witty lama) has published a report on his recent trip to India as the Wikimedia Foundation's fellow for GLAMs (collaboration with cultural institutions).
• Legal concerns about freedom of panorama and 2D reproductions in Germany: The German Wikimedia chapter recently published a legal opinion concerning the use of images on Wikipedia under German law, commissioned from its longtime counsel, Berlin law firm JBB. Regarding the reuse of reproduction photographs of two-dimensional public domain (gemeinfrei) works – the German equivalent of Bridgeman v. Corel -, it controversially stated that as long as the debate about the corresponding clause in German law (§ 72 UrhG) was still open, it had to be assumed that such reproductions "can not be used in Wikipedia unless the (reproduction) photographer has permitted the reuse". Another controversial conclusion was that while photos taken from a public place of copyrighted buildings and sculptures that are permanently located in such a place may be freely reproduced under the freedom of panorama clause (see also Commons:Freedom of panorama) as long as they are attributed, it is nevertheless not possible to publish them under a free license such as the CC-BY-SA, because the same clause (§ 59 UrhG) prohibited modifications. While presenting the expertise, Mathias Schindler from Wikimedia Deutschland emphasized that it "does not have an immediate impact on current practices, e.g. on Wikimedia Commons, and should be understood as a contribution to the continued discussion within the projects."
• Wikimedia Hungary report: The Hungarian Wikimedia chapter published its report for February 2011.
• Catalan report: The Amical Viquipèdia ("Friends of Wikipedia") association of Catalan Wikipedians has published its report for December to February.
• Wikimedia Netherlands report: Wikimedia Nederland published their report for February, mentioning among other things ongoing work on the upcoming Europe-wide "Wiki Loves Monuments" campaign
• New article feedback analysis: A further analysis on data from the pilot project exploring the use of the Article feedback tool has been published
• DMCA takedown: The Wikimedia Foundation recently received and acted upon a DMCA takedown notice by CEDRO, a Spanish author's rights society, concerning a link on the Spanish Wikipedia.
• More than 9000 articles deleted on the Hindi Wikipedia: In a blog post titled "Quantity over quality: The sad story of Hindi Wikipedia", Utkarshraj Atmaram (User:Utcursch) reported that over 9450 of its articles (nearly one seventh of all articles of the Hindi Wikipedia) appeared to be exactly identical except for one word in the lead ("XYZ is a village of Uttarakhand, India"), despite being over 5000 bytes long – the rest of each article was mostly filled by a brief history of the Uttarakhand state. The articles have since been deleted.

← Previous "News and notes"

Next "News and notes" →

In this issue

14 March 2011 (all comments)

News and notes

In the news

Features and admins

Arbitration report

Technology report

Editcountitis

Discuss this story

These comments are automatically transcluded from this article's talk page. To follow comments, add the page to your watchlist. If your comment has not appeared here, you can try purging the cache.

The relative drop in retention is obviously worrying. I am wondering though, do we have numbers available on the net amount of editors who are still active one year later? A quick look at the chart suggests (though there is no evidence) that this number is fairly constant. If it is, or is even rising slightly, there is less to worry about. Does anyone know if and where these numbers can be found? Martijn Hoekstra (talk) 09:15, 15 March 2011 (UTC)[reply]

I think the information you want is at Editor_Trends_Study/Results. Even if that's not what you're looking for, it is an interesting & informative read. -- llywrch (talk) 17:07, 15 March 2011 (UTC)[reply]

It was an interesting read. I'm wondering though where I could find the tabular data from these charts, if available. Martijn Hoekstra (talk) 21:31, 15 March 2011 (UTC)[reply]

As you can see at Editor Trends Study/Software, both the software used to generate the datasets from the Wikipedia XML dumps and the tools for analysis are open source. That page has step by step instructions for recreating his software platform if you want to run your own analysis -- more at Methodology too. My understanding is that it's actually not in tabular format currently, since he stores everything on his server hosting MongoDB (which is document-oriented). However, Howie who worked on the study might have it in CSV form though in order to generate the graphs, so you should ask on the talk page of the March Update. Steven Walling at work 22:57, 15 March 2011 (UTC)[reply]

I'll repeat here a suggestion I made at Wikipedia talk:Modelling Wikipedia extended growth, but that page doesn't appear to attract much attention any more. I'd be curious to know what the trend in the ratio of stubs:non-stub articles. If the ratio since the 2006 peak of new articles has been in favor of non-stub articles, that would support the "low-hanging fruit" hypothesis of declining new article creation -- viz., it is easier to improve an existing article than to create a new one. But if the ratio has remained roughly the same then the cause of the new article fall-off might be due to increased barriers to new article creation. (I don't know whether the XML data dumps would permit one to easily determine how the number of links to a page or template increased or decreased over time, which is the only way I can think of to arrive at this ratio.) -- llywrch (talk) 17:44, 16 March 2011 (UTC)[reply]

Article count inflation happens all the time with (what should be) the smaller Wikimedia wikis. Recently, the Malagasy Wiktionary jumped from 10,000 to 100,000 entries in the space of five months with entries copied verbatim from other wikis, such as "stop". – Minh Nguyễn (talk, contribs) 08:08, 16 March 2011 (UTC)[reply]

• Hats off to User:Utcursch for recognizing the value in providing useful quality content over junk stub farms. I applaud his decision. -- œ^™ 08:22, 16 March 2011 (UTC)[reply]

• The thing I don't get is why people who ought to know better are surprised that recent new editors don't stick around as long and blame it on how hard it is to use tools, that they aren't feeling welcomed, etc. Certain people are more suited for editing an encyclopedia than others. Individuals who read books, have editing experience and have a desire to share knowledge are only a subset of the general population. It seems to me that if someone hadn't already become an editor years back based upon hearing about Wikipedia and being interested, odds are good they just weren't all that interested. Early adopters are going to be more suited for the tasks involved, obviously, and the pool of available potential newbies will just get worse and worse as time goes on. That's not to say that we should not treat newbies with respect and welcome them, but I think we already do that to the extent we should be. As a basic philosophy we ought to do more to hang onto and encourage long time contributors who have proven themselves to the project. While Wikipedia is the encyclopedia anyone can edit, not everyone should be editing it regularly. They can contribute what they can add of value, and as time goes on and articles are already established newbies will have less available to do. And that's a good thing. DreamGuy (talk) 14:42, 16 March 2011 (UTC)[reply]

Wikipedia is getting big, is widely used and I would assume we could at least attract 10 times as many contributors. So why don't we? Two factors, learning to edit and motivation, of which I think the last is the most important. Most people are not self-starters and need to be told by someone they respect, community leaders etc. So even though we need to have a welcoming environment and keep on trying making it easier to contributer, I think we really need to have various big-shots to recommend contributing. Ulflarsen (talk) 19:22, 16 March 2011 (UTC)[reply]

Thanks for the idea. If the WMF acts on your idea, we'll likely end up seeing Jimmy Wales' mug at the top of every page for the rest of the year. And the PTB will also disable our ability to suppress these messages even for users with a login. -- llywrch (talk) 22:46, 16 March 2011 (UTC)[reply]