Attracting and retaining participants

The goal of increased participation

Some initial brainstorming suggests that several options (others here?) may be at play:

  • The lifecyle of a Wikipedia is limited: After a certain time period, a Wikipedia reaches a steady state, with fewer new contributors and slower growth in article creation.
    If this is the case, no goal around attracting new contributors may be necessary (the ongoing health of Wikipedias will rely on a smaller, but ever more active base of contributors who are able to maintain quality in existing articles and efficiently add articles in content gaps).
  • The lifecyle of a Wikipedia is not limited: Wikipedias could be redesigned to attract new participants or to engage existing participants in new ways:
    If this is the case, the goal may to create new ways to draw in new contributors--perhaps subject-matter experts, perhaps via new incentive structures to encourage participation.

 

Life-cycle of contributors to Wikipedia

The following analysis comes from Jose Felipe Ortega Soto - Wikipedia: A Quantitative Analysis.

Authors are most likely to leave early on in their Wikipedia experience

Contributors are highly likely to leave within a few days, and only 40% of logged users continue to contribute after 500 days.

 
The survival function of contributors to the top ten Wikipedias. Editors are highly likely to leave within their first few days, with fewer editors continuing activity as they spend more time on Wikipedia.

"Young" authors are highly likely to leave all the language Wikipedias, although this statistic is lower for the English and German language Wikipedias. This likelihood of leaving is highest at the start of a user's activity, and decreases rapidly over the first 15 days.

 
A "hazard" function for the top ten Wikipedias, showing the frequency at which users leave. The chance of leaving drops off quickly over the first 15 days of a user's activity.

Issues for core authors

Users who became "core authors", defined in this study as the 10% of contributors, did so in around 200 days. Once they reached the core, they generally stayed there for 200 to 400 days before leaving the core.

 
The red line represents the restricted average number of days it takes for a user to reach the top 10% of Wikipedia's contributors. The blue line represents the time spent as a "core author". Although it usually takes around 200 days to reach the core, the time spent in the core ranges from 200 to 400 days.

Contributions to featured articles come heavily from experienced users, with most contributions coming from authors with more than 300 days of experience.

 
In all language Wikipedias, the majority of contributors to featured articles have 300 to 1000 days of experience.

After leaving the "core" of authors, editors still tend to contribute. This statistic is similar for all language Wikipedias, but significantly lower for the English Wikipedia.

 
Survival time of editors after leaving the "core" of authors. More than 40% continue to contribute for at least another 500 days.

Barriers to increased participation

High reversion of edits for less experienced contributors

The Palo Alto Research Center (PARC) recently released a body of research [1] that suggests that new contributors to Wikipedia (those making 1 edit per month) face significant barriers to contribution. New contributors see 25% of their edits reverted, whereas the most active contributors see less than 2% of their edits reverted.

 
Less active users see their edits reverted at significantly higher rates than more active users.

Risk of new articles being tagged for deletion

In late 2009 WereSpielChequers conducted an experiment on the English Wikipedia to test how new users and their new articles are received by the community. He encouraged experienced contributors to create a new user account and create valid new articles. The results indicate that new users have a small but real risk of having valid articles deleted, with five out of 63 being deleted. But they have a substantial risk of feeling their article was being rejected, with 23 at least tagged for deletion - though reassuringly most of the incorrect tags were declined by admins or other patrollers.

 
Articles make up between 40% and 80% of all pages on the top 10 Wikipedias

Editing activity appears to be shifting away from article creation and maintenance, and towards growth areas such as article reviewing, bureaucracy and the reference desk. For the top 10 language Wikipedias, the total number of pages dedicated strictly to articles varies between 40% and 80%. [2]

This means that a large proportion of pages are now dedicated to purposes other than content creation or editing. These non-article pages include:

  • Article talk pages (for discussing improvements to articles)
  • User and related talk pages (for user profiles, draft articles and the like)
  • Wikipedia and Wikipedia talk pages (for reference queries, article reviewing and issues concerning Wikipedia, including norms/guidelines)
  • Image and related talk pages (concerning graphics on Wikipedia)
  • MediaWiki and related talk pages (concerning the software underpinning Wikipedia)
  • Template and related talk pages (concerning templates for easy content creation/linking)
  • Help and related talk pages (concerning efforts to offer assistance in using Wikipedia)
  • Category and related talk pages (concerning the categorization of pages)

Even if the pages tangentially related to content creation (e.g., image pages or category pages) are excluded, a significant proportion of pages on Wikipedia are dedicated to other purposes. Further research is needed to establish how much of this is:

  1. The bureaucracy of the project,--including guidelines, protocol, manual of style, dispute resolution, appointment of admins and other functionaries.
  2. Quality improvement, including guidelines, manual of style, article reviewing and software discussion.
  3. Answering queries from readers, mainly at the reference desk or article creation.
  4. Editor training, such as at the helpdesk

Technical and comfort barriers

The following findings come from the preliminary findings from the 2009 survey of 100K+ Wikipedia users. [3] Bridgespan has done some additional analysis on top of the published findings.

Surveyed non-contributors cite two key types of barriers to participation: perception that contribution is not valuable and technical barriers.

 
Non-contributors face barriers to participation.

Surveyed non-contributors cite two factors that might increase participation: feeling more valued and facing fewer technical barriers.

 
Factors to increase participation link to feelings of value and technical barriers.

Some caveats here: It is unclear how representative this sample is of the Wikipedia community at large. For example:

  • This survey had an over-representation of editors vis-a-vis the general population.
  • There was an over-representation of Russian respondents, who were removed from the study pending some explanation for this anomaly.

Other community health issues

The lack of women on Wikipedia may be indicative of a collaborative style that is less welcoming to women. This lack of women appears to have affected the content created on the Wikimedia projects. To name one example, of 174 women involved with the Nation Women's History Project, Wikipedia appears to be missing articles on 23 of them. [4]

More generally, it also appears that the community lacks strong incentive structures to reward high-quality contributions.

Topical areas that non-participants might be interested in and that are missing from Wikipedia

See Data on content breadth and composition for current content areas covered.

Any insights on what's missing?

Enablers of increased participation

A study from Jose Felipe Ortega Soto - Wikipedia: A Quantitative Analysis found an interesting correlation. Editors who contribute to talk pages are likely to leave later into their experience, as are editors who contribute to featured articles. Editors who contribute to both featured articles and talk pages are much more likely to stay at Wikipedia for a longer time. However, the study cautions that this is only a correlation.

 
The survivability of editors over time on several Wikipedias. The blue line shows that editors who contribute to talk pages and featured articles are most likely to remain an active contributor. The black line represents the survivability of Wikipedians who did not contribute to featured articles or talk pages. The red line shows editors who contributed to talk pages, and the green line shows editors who contributed to featured articles.
















.

Under-represented groups among contributors

See Demographics of Wikipedia contributors for demographics of current contributors.

Underrepresented demographics include:

Ways that Wikimedia could attract new participants or retain existing participants

Create new opportunities and roles for active participants

  • Frequent contributor transitions into role of in-person outreach/training
  • Frequent contributor shifts into role with chapter
  • Etc.

Add new functionality to Wikipedia

  • Make it easier to track changes to articles of interest and thus ensure contributions survive.
    • by default set new users to "watch" articles they create.
    • point first-time editors to a tutorial on what the watchlist does.
    • enable email notifications of page changes on wikipedia (currently only the commons and meta sites have it; explanation for why is that too many messages would be sent, but bandwidth is also used for people clicking through to see what changes have been made)
      • Always show in preferences where the "enable notifications" box would be in preferences, with a note explaining that it is not available in wikipedia (until it is)
      • By default, do not include minor page changes in email notifications
      • Include diffs in email notification about changes
      • Provide option for email notification if articles watch are proposed for deletion separately from/in addition to "changes" notification
      • Create a summary or "digest" option for email that can digest change notifications by day or by week
  • Add social networking elements to MediaWiki
    • provide tools for users to rate, recommend/review, and "share" articles
      • might want to differentiate between rating quality of article and importance of topic
    • Provide ways for contributors to "watch" page creation or editing by other editors of interest
    • provide ways to indicate social connections between editors.
    • raise the visibility of groups and projects to new users.
    • increase the social interactivity of being part of a project
  • Draw contributors back in to continue refining articles by allowing readers to ask questions they feel haven't been adequately answered in an article.
    • "Reward" contributors that answer reader questions.
  • Allow contributors to subscribe to feeds by "topic area" so they can see what questions readers are asking.
  • Allow contributors to subscribe to feeds so they can see which terms have been "turned red" on specific pages.
  • Make it easier to add verifiable information. (More user-friendly citation tools, let people add links to research that others can summarize, etc.)
    • Create a Wikipedia toolbar that lets you flag news articles, and categorize them by Wikiproject/topic. This will make it easier to get verifiable information onto Wikipedia.
  • Discourage deletion of quality content
    • Add functionality that lets editors "mark" text that they have verified and approve of, so that text gets more gravity and stability in an article
    • "penalize" contributors and admins who demonstrate a pattern of removing others' verified work
    • make it easy for users to "report" a user or IP address that is deleting content to admins

Introduction of more social elements to user interface

At present, the Wikipedia interface has limitations in its ability to facilitate social interactions. For example, it is difficult to identify the core group of contributors who created most of the content in an article without extensive history browsing. As another example, MediaWiki (the software underpinning Wikipedia) is being used for both article creation and discussion, with some blog software offering discussion functionality that MedaWiki lacks.

Revamping the interface, with a focus on the social element of content creation, could have far-ranging effects. Introducing a more social atmosphere to Wikipedia could consist of several components:

  • Users are granted space for pictures/avatars, thereby creating a visual representation of themselves
  • Every article page would have a list of the top 3-5 contributors to the page, with pictures/avatars posted along with a link to the contributors’ user pages
  • Users would be able to join topical groups, based on their editing interests (e.g., “18th century American history)
  • Users would also be able to track the editing patterns of their friends, a la Facebook or Twitter (e.g., “Susie just changed an article on the Tyrannosaurs rex. Click here to see what she changed”)

Such an approach would make it easier for new contributors to:

  • Quickly identify how best to contribute
  • Quickly form social connections (via groups and “following”)
  • Feel more comfortable in a new setting, with features similar to social networking sites

This would likely require a significant software revamp and might also require the introduction of WYSIWYG (what you see is what you get) editing software, which eliminates the need to be familiar with Wiki syntax.

Potential partnerships

Potential options include:

  • Social networking sites
    • LinkedIn - particularly if people cite "professional development" as a reason for contributing to Wikipedia
  • Libraries and other cultural institutions holding workshops with visitors
  • Engaging high school teachers: contributing to Wikipedia articles may be an excellent way to teach students basic research and summarization skills
  • International Organizations working with local populations in developing context appropriate educational materials
  • Other?

Examples of other mass collaboration portals successfully reaching out to new segments

What has happened with:

  • Facebook?
  • Epicurious?
  • Yelp?
  • Etc.

Implications of expanding core of most active users vs. expanding group of occasional contributors

Given that 10% of contributors contribute ~90% of edits on Wikipedia (see Breakdown of editing patterns), it appears that a core group of contributors drives most content changes on Wikipedia.

Other potential/emerging ideas

Some potential ideas:

  • Professor reviews 50+ articles, sends comments back to Wikipedia community
  • Additional outreach to content "specialists" (e.g., cancer researchers at Stanford)
  • Allow readers to ask questions if an article doesn't have what they're looking for, on the english wikipedia this is done at en:Wikipedia:Reference desk
  • Perhaps the model followed by the Encyclopedia of Earth at www.eoearth.org could be considered
  • Allow readers to request that an article be written, on the english wikipedia this is done at en:Wikipedia:Requested articles
  • Welcome all new users who've contributed something that hasn't been reverted as vandalism. On commons this is done by a bot
  • Other?

Notes

  1. "More details of changing editor resistance in Wikipedia." PARC. 7 August 2009. http://asc-parc.blogspot.com/2009/08/part-2-more-details-of-changing-editor.html
  2. "Wikipedia: A quantitative analysis." Jose Felipe Ortega Soto. 2009. http://libresoft.es/Members/jfelipe/thesis-wkp-quantanalysis
  3. Wikipedia Survey – First Results. April 2009. http://upload.wikimedia.org/wikipedia/foundation/a/a7/Wikipedia_General_Survey-Overview_0.3.9.pdf
  4. "Gender Bias in Wikipedia Coverage?" Joseph Reagle. http://reagle.org/joseph/blog/social/wikipedia/wp-eb-gender-bias-coverage.html