Editor Survey 2011/Executive Summary
With approximately 81,000 active editors1 in a month2, Wikipedia is one of the most successful examples in online collaboration in the history of the Internet. But precious little is known about Wikipedia editors. Who are they? What motivates them? What are their experiences with contributing to Wikipedia? To answer some of these questions, we conducted the first ever semi-annual survey of Wikipedia editors in April 2011. The survey was conducted on Wikipedia and presented to logged-in users. The study focused on the following key research areas as linked to our strategic plan:
- 1. Contribution for sharing knowledge freely WITH every single person on the planet: The Wikimedia movement, since the very beginning, has been grounded in the mission of sharing free knowledge with every single human being. The survey shows that our editing community is highly aligned with the core mission of the movement. Editors pointed out that volunteerism to share knowledge is the number one reason for contributing to Wikipedia. The contributors, even through all of the growth and changes over the years, continue to share this fundamental raison d'etre of Wikipedia. A "Wikipedian", whether or not she or her sees him or herself as part of any given "community" (and there are good reasons not to, or to see projects as existing at the juncture of many communities), agrees to respect some basic editorial rules. Even those who believe that " an encyclopedia is a market in theories and facts" or a "battlefield of ideas" can accept rules designed to make the markets or battles fair and permit new participants in. Given the shared goal of usefulness to literally everyone, it's relatively easy to justify rules that make pages more useful to readers even if this inconveniences writers (a rule understood on some level by all wiki editors). This distinguishes Wikipedia, and wiki, from other social media and could and should be emphasized in any effort to attract new editors from new communities, especially those who were previously hostile to Wikipedia.
- 2. Editing tools and infrastructure improvements: Over the years, Wikipedia editors have created many tools and policies both for editing and performing the day-to-day administrative tasks required to manage huge knowledge resources and activities on the world's fifth largest website. The survey found that there is a general agreement that both the community and WMF can – and should – improve our technology infrastructure and develop new features to make editing and administration easier. Or, rather, as easy as it can be without making the final results less useful to the reader or non-editing user. WMF has already started work on improvements and will accelerate work alongside the community in the coming months. WMF is also undertaking research work that will help the editor community assess opportunities for improvement of tools and practices. As the survey indicates, interactions among editors are key drivers for a better editing experience. Features like WikiLove provide positive reinforcement and fun, and mirror the mechanisms available in other social media such as facebook.
- 1. Demographic characteristics of editors: Since we haven't had good demographic data about Wikipedia editors, a caricatured profile of Wikipedia editors has emerged over time: a male graduate student who programs, supports open source, plays massively multiplayer online games, and lives in the US or Europe. Concerns about systemic biases introduced or propagated by relying on too-narrow a base of editors is an old debate in Wikipedia, dating back at least to debates between Larry Sanger and 24 which contributed to Sanger's resignation, and to more organized attempts to counter bias. As of 2011, Wikipedia still struggles with systemic bias and cliques, sometimes even using non-objective language (like "troll") in its official policy statements. However, representativeness of editors has improved: The 2011 survey shows the caricature is a bit broader than it was in 2002: According to the data, if there is a typical Wikipedia editor, he has a college degree, is 30-years-old, is computer savvy but not necessarily a programmer, doesn't actually spend much time playing games, and lives in the US or Europe. But “typical” doesn't tell the whole story. Users have many varied backgrounds - all things considered, the person behind the username is quite likely different from you. To add greater amount of diversity to the mix requires sensitivity to gender, geography, culture, communication styles and other differences. Including whether or not editors wish to see themselves as part of a "community" or participating in a more neutral and objective process which simply seeks to be useful to its readers.
- 1. Increasing gender diversity: Our editing community continues to suffer from a lack of women editors. The survey provided an even starker view of this than previous studies (only 8.5% of editors are women). It is a strategic priority to address this imbalance. The survey did find that the total percentage of women Wikipedia editors has increased somewhat in the last few years, but we still have a lot of ground to cover. We can attract women editors partly by introducing tools and features that make editing simple for everyone, though especially for women, since our women editors are less likely to code and program. We have also seen great successes in the participation of women via our Wikipedia in the classroom initiatives. These efforts that are expanding around the world tend to bring in a more representative proportion of men and women contributors. Our survey found that positive feedback is a driver for overall editor satisfaction, and so is the nature of the interactions they have with other users. Social approaches (such as experiments to assign buddies to new editors) may also be more effective with women than men.
- 2. Harassment: Contrary to the perception of some, our data shows that very few women editors feel like they have been harassed; Very few feel that Wikipedia is a sexualized environment. That said, all users should ensure that all female Wikipedians have positive interactions, and that no one is harassed due to their gender, sexuality, ethnicity or nationality. It would also be helpful to reduce the use of legal or loaded terms such as "stalker" to describe petty disagreements and reserve it for actual cases of harassment that causes emotional distress or clearly implies physical risks; A female user facing a genuine "stalker" does not need a clique of self-appointed counter-harassers acting on their own authority or beliefs of what constitutes "stalking", she needs a genuine investigation and response (possibly via ArbCom) that is consistent and effective at discouraging abuse.
- 1. Editor decline and the community: The Wikimedia movement has made increasing its editor base to 200,000 by 2015 a major priority. But the recently concluded Editor Trends Study discovered an alarming trend of flattening participation across all language projects. Looking closely at English Wikipedia, the study found significant decline in editors with more than 10 edits a month. It has been hypothesized that edit wars, reverts and acrimony among editors is a contributor to this decline. We found that, overall, editors have a very positive opinion of their peers, but many reported negative interactions and what they perceive as harassment by others. In addition, negative interactions reduce the likelihood of editing in the future. On the other hand, positive interactions, like helping others in editing, and peer recognition, not only make editors have a more positive opinion of the community, but increase the likelihood of editing in the future. Expectations of "community" may play some role in some users' disappointment, as most of the WMF's own official statements freely throw around terms like "community" without acknowledging that many Wikipedia editors provide valuable content without much (or any) concern for the social processes around that content. Academic sticklers and political drum-beaters often provide good references and much-needed clarity and precision, but relatively few people can deal with them on equal terms. As the number of open topics decreases and the scrutiny of controversial topics increases, this intensity and conflict with users who perceive themselves as having a stake in the content will increase. In other words, as Wikipedia becomes more influential, influence on it becomes a higher-stakes game that attracts more intense (and perhaps some ruthless) players.
- 2. Positive reinforcement: Acknowledging the effort of editors is important to reverse the editor decline. It is a commonly held view that editors just want to see their articles improve and read by lots of people and they don't care about the opinion of their peers. Most users apparently do not agree: the survey finds that acknowledgement of peers via a nice note or a barnstar (or kitten) is valued even more highly than achieving featured article status. To provide each other with positive feedback and create tools to make that easier should accordingly remain a high priority.
- 3. Negative and positive interactions: It is important to encourage new mechanisms for celebrating and rewarding excellent editor contributions within the community. Clearly, maintaining a positive environment on wikis through positive feedback loops is essential both for supporting current editors and attracting new ones. The Wikimedia movement will continue to support current editors through experimental tools like WikiLove that provide a channel for positive and enjoyable interactions within the community. The movement will also support the recruitment and acculturation of new editors by encouraging a welcoming environment on Wikimedia projects. But we would also like to call upon our community to provide positive feedback to others. Negative experiences matter. Editors don't have hearts made of stone. Reverts without an explanation (which as a rule are almost never justified and should cause immediate scrutiny on editors that do so) can negatively define an editor’s experience on Wikipedia and make them less likely to continue editing. However, editors are here to learn and improve. A revert with an explanation, providing at least some path to a new and better interaction, has no negative impact on an editor's desire to continue contributing and is, in fact, seen as a positive interaction. We need to restrict unexplained reverts, find ways to reduce negative experiences and refine our automated tools to do a better job of differentiating a good faith edit from deliberate vandalism.
- 4. Identifying and reducing clique influence: While perceived fairness of edit and article inclusion and exclusion has generally increased over Wikipedia's history, it remains a common practice to claim that any reason for exclusion constitutes clique influence and to seek publicity for articles 'deleted' or 'censored' from Wikipedia . Some of these claims are or may be valid. Where the excluded material is specific to women's rights or causes or champions (notability thresholds for feminist activists, for instance) the practice should generally be to err on the side of inclusion.
Locations and languages
- 1. Increasing geographic diversity: The data from the survey shows that the majority of Wikipedians hail from North America or Europe, and to meet our strategic goal of increasing diversity – specifically, attracting more editors from the Global South – we will continue in our efforts to expand Wikipedia’s global footprint. We have a growing number of chapters around the world and WMF has set up an office in India and hired staff to support growth in India. Similar efforts are also beginning in Brazil. Our goal is to support the growth of the editing community with accelerated growth in areas where we are not yet strong.
- 2. Mobile and the Global South: In some of these regions, like India and Africa, desktop Internet has yet to see broader penetration, though mobile Internet is expanding rapidly, and it is no surprise that the mobile phone is the most popular device among editors. When Wikimedia's projects reach the three billionth user projected in the early (2002) debates on it's direction, they will likely be reached predominantly by low end phones by poor people. WMF has made it our priority to increase mobile page views, and we are currently revamping our mobile platform to provide better and faster access to smartphones as well as feature phones that don’t typically have apps or can’t be synced with computers. The new platform will have in-built editing functionalities that would allow for paragraph edits, sentence edits and picture uploads to Wikimedia Commons. Lastly, we are looking to establish partnerships with network providers in key strategic geographies like India and Brazil to provide access to Wikipedia at zero or near zero cost. This would help us increase our reach and bring free knowledge to those who can’t afford to pay for data access.
- 3. Language diversity: Survey respondents edit Wikipedia in more than 100 languages. Interestingly, only 38% edit primarily in English, but 76% of all editors edit the English Wikipedia. This is quite shocking and points to the fact that our editor community does not reach far into non-English proficient communities. This poses a challenge for us as we seek to move into new geographies and segments of the population where English literacy is not prevalent. As we work towards increasing our global footprint we are committed to supporting less mature language projects and ensuring quality articles in native languages, especially in the Global South. We believe global partnerships with universities, cultural institutions and other groups who are aligned with our mission will help us create quality content in native languages. In India, as part of the Global Education Program, while most of the students are working on the English Wikipedia, some of the newly appointed campus ambassadors are focused on the Marathi (the native language in the state of Maharashtra) Wikipedia.
- 4. Digital divide: Users who lack unlimited data or free Wi-Fi access, especially for mobile devices, are particularly disadvantaged and inhibit (and should inhibit) deployment of data-rate-intensive features. A desirable solution to this problem that should be investigated is convincing mobile service providers to include all WMF projects in their lists of "free access" or "unlimited access" services alongside facebook, email and other essential applications. Thus, regardless of local data rates, it would be possible to freely access Wikipedia et al without fear of getting a big bill. Another solution would be to support cacheing when Wi-Fi access is available to minimize the use of 3G/4G bandwidth (and thus the charges). A recruiting effort to find disadvantaged editors to test such solutions and to propose new ways to radically reduce the cost of accessing WMF projects is probably justified, especially as the "typical" user has no such problem.
Technology and networking
- 1. Social networking: Our data shows that Wikipedia editors frequently use social networking technologies like Facebook. The Wikimedia movement has not systematically used social networking platforms as a means of communication, and there has been resistance on philosophical grounds to integration with social networks. The fact that most editors regularly use Facebook, and a significant minority use Twitter, is notable. We know that women tend to be more active on social networks than men3 and social networks are growing rapidly in priority countries such as India and Brazil4. These facts raise the question of how these pervasive tools might support editor interaction and communications within the movement. The case for leveraging social networking may be important to the growth of the Wikimedia movement.
- 2. Mobile technology: While a higher than average percentage of all users have access to smartphones (Android, iPhone, BlackBerry), the average user is far less likely, and given the growth of usage in the global south, will get much less likely in future, to use more than simple SMS messaging. Accordingly surveys of readers not editors may be the way to determine the mobile strategy. Efforts to make mobile editing possible may not, given the editor profile, increase representativeness or readership of mobile users. In this respect in particular, the person writing is not the person reading.
Chapters, board, and functions
- 1. Raising awareness: Most editors are not knowledgeable about chapters or board elections. Since the Wikimedia movement is decentralized, the voices and opinions of community members are pertinent, and the WMF will continue to experiment with new ways of soliciting participation from editors in board elections. There is a need for chapters to perform outreach within their countries to inform and involve community members in chapter work.
- 2. Support for work of the WMF and other volunteers: While editors were humble in their assessment of their own activities, they valued the work of their peers in the movement as well as the efforts of the Wikimedia Foundation. The WMF was glad to see that editors who sought information about us were generally satisfied, though we continue to find ways to improve.
- 3. Toning down exclusionary and clique-encouraging rhetoric: The use of some common words or practices should be utterly excluded from WMF's official discourse:
- "troll" - this term has no objective definition and is usually used simply to label one's debate opponent as hostile, disruptive or having some sort of "behavior" problem, ignoring or downplaying the substance of their arguments
- any words that imply that one editor has insight into someone else's motive.
Other words should be sharply restricted and used only with extreme due care:
- "stalker", "threat", "attack", "assault" – these terms should be used only in their legally defined sense and treated with due seriousness; the use of any of these terms as a rhetorical device to describe insults or any annoyingly persistent action should be sharply discoursed including by direct sanction.
- "community" – most real world abuses historically have been carried out by persons who claimed (often with reason) to serve "the community" (defined as a single monolithic entity, sometimes derived from a majority view or just a majority of powerful actors). It's long past time WMF acknowledged that as the projects expand and reach more people, they will be less and less likely to perceive themselves as a single "community" and more likely to perceive a WMF project as a neutral negotiating (or fighting) ground for their disputes and conflicts. WMF should accordingly strive to present itself as less of a "community" and more of a fair and judicially minded process that encourages political virtues (liveliness, adaptability, compromise, humor, etc.) and is open to anyone with a good point, even if they offend or annoy someone else.
In other words, Wikipedia must grow up and acknowledge its work as more diversified, political and contentious than that of any other social medium, and that this is exactly why its products and services are more important. It should abandon the "single community" ideal and acknowledge that it will be only one small factor in global harmony, albeit an extremely central one. The solution to editorial representativeness and reduced systemic bias, and increased participation, is to acknowledge the ugly reality of a world riven by conflicts, cliques and factions, and to present a process to find some version of the truth that, while not perfect, is at least more accessible and fair than the UN, civil courts, mass media or other venues. Not a "place" where one must "join" a "community" to "belong", but rather a means by which persons who see themselves as part of many communities, can find just enough common ground to acknowledge some undisputed basic reality.
- Executive Summary
- Editing Activities
- Women Editors
- Wikipedia Community
- Location & Language
- Technology & Networking
- Foundation & Chapters
1. An active editor makes 5 or more edits in a month
2. As of Jun 2011: stats:EN/TablesWikipediaZZ.htm#activitylevels
3. Women spend 30% more time on social web than men http://socialtimes.com/women-more-on-social-web-than-men_b18934
4. BCG report: The Internet’s New Billion http://www.slideshare.net/agarwalvaibhav/bcg-internet-report
5. countering systemic bias