The Signpost

Recent research

Reciprocity and reputation motivate contributions to Wikipedia; indigenous knowledge and "cultural imperialism"; how PR people see Wikipedia

A monthly overview of recent academic research about Wikipedia and other Wikimedia projects, also published as the Wikimedia Research Newsletter.

What drives people to contribute to Wikipedia? Experiment suggests reciprocity and social image motivations

Wikipedia works on the efforts of unpaid volunteers who choose to donate their time to advance the cause of free knowledge. This phenomenon, as trivial as it may sound to those acquainted with Wikipedia inner workings, has always puzzled economists and social scientists alike, in that standard Economic theory would not predict that such enterprises (and any other community of peer production, for example open source software) would thrive without any form of remuneration. The flip-side of direct remuneration – passion, enthusiasm, belief in free knowledge, in short, intrinsic motivations – could not alone (at least as standard theory goes) convincingly explain such prolonged efforts, given essentially away for free.

Early on the dawn of the Open Source/Libre software movement, some economists noted that successfully contributing to high-profile projects like Linux or Apache may translate in a strong résumé for a software developer, and proposed, as a way to reconcile traditional economic theory with reality, that whereas other forms of extrinsic motivation are available, sustained contribution to a peer production system could happen. But what about Wikipedia? The career incentive is largely absent in the case of the Free Encyclopedia, and is it really the case that intrinsic motivation such as pure altruism cannot be really behind the prolonged efforts of its contributors?

To understand this, a group of researchers at Sciences Po, Harvard Law School, and University of Strasbourg (among others) designed a series of online experiments with the intent of measuring social preferences, and administered them to a group of volunteer Wikipedia editors to understand whether contribution to Wikipedia can be explained by any of the main hypotheses that economists have thus far formulated regarding contribution to public goods.[1][2] The researchers considered three hypotheses, two for intrinsic and one for extrinsic forms of motivation: pure altruism, reciprocity, and social image motives.

In more detail, the researchers asked a number of Wikipedia editors and contributors (all with a registered account) to participate in a series of experimental games specifically designed to measure the extent to which people behave according to one or more of the above social preferences – for example by either free-riding or contributing to the common pool in a public goods game. In addition to this, as a proxy measure for the “social image” hypothesis, they checked whether participants ever received a barnstar on their talk pages and whether they ever chose to display any of these on their user page (coding these individuals as “social signallers”). Finally, they matched each participant with their history of contribution of the participants, and sought to understand which of these measures can explain their edit counts.

The results suggest that reciprocity seems to be the driver of contribution for less experienced editors, whereas reputation (social image) seems to better explain the activity of the more seasoned editors, though, as the authors acknowledge, the goodness of fit of the regression estimates is not great. The study was at the center of a heated debate within the community about the usage of site-wide banners for recruitment purposes. On December 3, one of the authors gave a presentation about the results at Harvard, which is available online as an audio and video recording. According to the Harvard Crimson, he remarked "that the study is still in progress and more data needs to be collected". The results are so far available in the form of a conference paper and as an unpublished working paper.

Does "cultural imperialism" prevent the incorporation of indigenous knowledge on Wikipedia?

The "People are Knowledge" documentary video
Nuxalk masks

A draft chapter[3] of a book to be published in early 2014 presents the issue of incorporating into Wikipedia "Indigenous Knowledge" (IK) – human knowledge that is not a part of the codified and peer-reviewed Western-style publishing, but is rather transmitted orally in other parts of the world. The problem is not new; perhaps most notably, it was described in the 2011 documentary "People are Knowledge", which was produced by Indian Wikimedian Achal Prabhala and funded by the Wikimedia Foundation as a fellowship project. The general problem is that Wikipedia relies on written reliable sources for verifying its materials. This article describes Wikipedia's policies and editing practices that are relevant to the problem of incorporating Indigenous Knowledge. In describing these it makes a rather problematic claim — that "the 'currency' of Wikipedia is edit count". Many Wikipedia editors will find this claim wrong and even offensive, as quality, rather than quantity, counts for an editor's reputation, and in any case the content is more important than the creator.

The article presents several valuable and thought-provoking examples of how the rigid referencing rules of the English Wikipedia go to extremes and do not necessarily reach the goals of ensuring notability, verifiability and reliability. It notes, for example, that because many of Wikipedia's editors are laymen who want to work quickly and fill the gaps that interest them, they are likely to cite sources partially without reading them completely and deeply — thus undermining the sources' reliability. Another example is Gi-Dee-Thlo-Ah-Ee, a Cherokee woman who was the subject of a book that was included in the Library of Congress. An article about her was deleted from the English Wikipedia, the main reason being that the book was not deemed an independent reliable source, because it was published by the Cherokee Nation. The case of the article Makmende ("the first Kenyan Internet meme") is also cited, although the validity of this example has been questioned (Signpost coverage: "Essay examines systemic bias toward African topics, using disputed deletion example").

This work on oral citations by Achal Prabhala, as well as Prabhala's practical attempts to challenge the English Wikipedia's citation policy is the subject of a large part of this article. It shows that until now Prabhala's attempts have mostly failed, because the editor community found his citation practices unacceptable. The article analyzes the typical responses of the people who are opposed to oral citations and shows some problems with them. However, it doesn't yet give any useful resolution to the issue and labels the opposition to oral citations as "cultural imperialism".

Despite its shortcomings, this article is a good presentation of the issues at hand, as well as of their importance, and it is a good summary of the work done in the field until now.

The structure of the Traditional knowledge article.
Pronunciation of Skwxwu7mesh (Squamish).

The second author, Maja van der Velden, had published another article on the same subject some months ago,[4] also referring to Achal Prabhala's oral citation project and comparing it to two other initiatives, Text, Audio, Movies, and Images (TAMI), and the Brian Deer Classification (BDC). TAMI is a database on Australian Aboriginal culture, BDC a library classification system in use at the Xwi7xwa Library, which specializes in Canadian Aboriginal culture. Indigenous communities were involved in the design of both, resulting in some marked differences from Western (and Wikipedia) design habits, e.g. a flat hierarchy of only four categories in TAMI (those represented in the acronym), or the lack of a "Canada" class in BDC ("United States" exists, at the same level as "Maoris").

She criticizes the merger of the indigenous knowledge entry on the English Wikipedia into traditional knowledge on principle grounds, adds that the merger did not actually merge content from the former into the latter, and takes issue with the focus of the traditional knowledge entry being so much on intellectual property, to the point that she added a screenshot of the article's table of contents (much like the one pictured here).

A stop sign in Mi'kmaq

After outlining how matters of design are handled on Wikipedia, van der Velden discusses whether fulfilling its mission of providing access to the sum of human knowledge might benefit from decentralizing design decisions, which brings her to the regularly recurring ideas of decentralizing Wikipedia and to a discussion of interwiki links that manages not to mention Wikidata.

Overall, the article is an interesting and in parts thought-provoking contribution to the activities around increasing diversity within the Wikimedia community (see, for instance, the Wikimedia Diversity Conference, held in Berlin earlier this month). It would have benefited from a more detailed description of TAMI and BDC and from suggestions as to how their respective community engagement experiences could be transferred and adapted to cross-cultural collaboration in Wikimedia projects.

In light of the recent increase in for-hire editing on Wikipedia, often carried out by PR professionals, another timely study has been released,[5] a survey among PR professionals, as a followup to one covered in the April 2012 edition of this research report ("Wikipedia in the eyes of PR professionals"). The surveys examine how familiar the PR professionals (working not only for for-profit organizations, but also for non-profits, educational institutions, government institutions, and others) are with Wikipedia rules. 74% of respondents noted that their institution had a Wikipedia article, a significant (5%) increase over the 2012 survey, though over 50% of the PR professionals do not monitor those articles more often than on a quarterly basis. The study confirms that there is a steady but slow increase in PR professionals who have made direct edits to Wikipedia; 40% of the 2013 survey respondents had engaged with Wikipedia through editing (with about a quarter of the respondents editing talk pages, and the remainder directly editing the main space content), compared to 35% of the 2012 survey respondents. Over 60% agree that "editing Wikipedia for a client or company is a common practice", a slight but statistically significant decrease from 2012. While "posing as someone else to make changes in Wikipedia" is not seen as a common practice, it is nonetheless supported by ~15% of respondents in the US and almost 30% elsewhere (though the latter number should be taken tentatively, as 97% of the survey respondents came from the US).

At the same time, approximately two thirds of the respondents do not know of or understand Wikipedia rules on COI/PR and related topics (defined in this study as Wales' 2012 "Bright Line" policy proposal, linked to his comment in a Corporate Representatives for Ethical Wikipedia Engagement (CREWE) from January 10, 2012, 5:56 am (accessible here – Facebook login required). Of those who had experience editing Wikipedia directly, thus breaking the rule, over a third (36%) did so knowing about it, thus knowingly violating the site's policy.

The significant breadth of ignorance about Wikipedia rules reinforces the point that even a decade after Wikipedia's creation, most of its users do not even realize that it is a project "anyone can edit", much less what it means: 71% respondents replied that they simply "don't know" "How Wikipedia articles about their clients or companies are started", which presumably indicates that they do not understand the basic function and capabilities of the article history function. A majority of other respondents (24% total) admit to writing it themselves; 3% hired a PR firm specializing in this task, 1% hired a "Wikipedia firm" (a concept unfortunately not defined in the article), and only 2% note that they "made a request through Request Article Page"). When it comes to existing articles, only 21% of the respondents wait for the public; the vast majority of the rest make edits themselves, with 5% outsourcing this to a specialized PR or "Wikipedia firm".

Respondents who had directly edited Wikipedia for their company or client said their edits typically “stick” most of the time. Over three quarters noted that their changes stick half the time or more often; only 8% said they never stick, always being reverted. This raises the question about the efficiency of Wikipedia COI-detection practices, as well as of their desirability (are we not reverting those changes because we don't realize they are COI-based, or are they reviewed and left alone as net-positive edits?).

60% of the respondents note that the articles about their clients or companies have factual errors they would like to correct; many observed that potentially reputation-harming errors last for many months, or even years. This statistic poses an interesting question about Wikipedia responsibility to the world: by denying PR people the ability to correct such errors, aren't we hurting our own mission?

The majority of respondents were not satisfied with existing Wikipedia rules, feeling that the community treats PR professionals unfairly, denying them equal rights in participation; even out of the respondents who tried to follow Wikipedia policies and who raised concerns on the article's talk page rather than directly editing them, 10% noted that they had to wait weeks to get any response, and 13% said they never received a response.

Regarding to the new editors' experience, it is also interesting to note that only a quarter of PR professionals felt that making edits was easy; the majority complained that editing Wikipedia is time-consuming or even "nearly impossible".

Accessing Wikidata data, a tutorial by Max Klein.

Report from the inaugural L2 Wiki Research Hackathon

A visualization of the Turkish Wikipedia community, with clustered sub-communities in different colors, generated using Gephi
Visualizing Wikipedia as a graph using Gephi, a tutorial by Haitham Shammaa.

On November 9, 2013, a group of Wikimedia Foundation researchers, academics and community members hosted the inaugural Labs2 Wiki Research Hackathon: the first in a series of global events meant to "facilitate problem solving, discovery and innovation with the use of open data and open source tools" (read the full announcement from the Wikimedia Blog). The event brought together attendees from local meetups in Oxford, Mannheim, Chicago, Minneapolis, Seattle and San Francisco and a number of remote participants. Participants began the groundwork around new projects studying Wikipedia including a study of newcomer retention focused on females, explorations of using Wikipedia as a multilingual corpus, an examination the effectiveness of helpdesks on Wikipedia and several others. A series of presentations were given and streamed during the event, including:

The organizers are planning to host a new hackathon in Spring 2014 and are actively seeking volunteers to host local and virtual meetups. (wrh@wikimedia.org)

Briefly

References

  1. ^ Yann Algan, Yochai Benkler, Mayo Fuster Morell, Jérôme Hergueux: Cooperation in a Peer Production Economy. Experimental Evidence from Wikipedia. Working paper, PDF.
  2. ^ Yann Algan, Yochai Benkler, Mayo Fuster Morell, Jérôme Hergueux: Cooperation in a Peer Production Economy Experimental Evidence from Wikipedia. Aix-Marseille School of Economics, 12th journées Louis-André Gérard-Varet (June 2013) [http://www.afse-lagv.com/lagv/submissions/index.php/LAGV2013/LAGV12/paper/view/185 �ŽPDF]
  3. ^ Peter Gallert, Maja van der Velden: "Reliable Sources for Indigenous Knowledge: Dissecting Wikipedia’s Catch–22". Draft, to be published in early 2014 as a chapter of the post-conference book for the Indigenous Knowledge Technology Conference (IKTC) 2011 (editors: N. Bidwell and H. Winschiers–Theophilus) PDF
  4. ^ Van Der Velden, M. (2013). "Decentering Design: Wikipedia and Indigenous Knowledge". International Journal of Human-Computer Interaction. 29 (4): 308–316. doi:10.1080/10447318.2013.765768. Closed access icon
  5. ^ Marcia W. DiStaso: Perceptions of Wikipedia by Public Relations Professionals: A Comparison of 2012 and 2013 Surveys. Public Relations Journal Vol. 7, No. 3, ISSN 1942-4604, Public Relations Society of America, 2013 PDF
  6. ^ a b Aaron Shaw, Benjamin Mako Hill: Laboratories of Oligarchy? How The Iron Law Extends to Peer Production (draft paper) PDF
  7. ^ Tolomei, Gabriele; Orlando, Salvatore; Ceccarelli, Diego; Lucchese, Claudio (2013). "Twitter anticipates bursts of requests for Wikipedia articles" (PDF): 5–8. doi:10.1145/2513577.2538768. {{cite journal}}: Cite journal requires |journal= (help)
  8. ^ Han-Teng Liao (2013) How does localization influence online visibility of user-generated encyclopedias? A study on Chinese-language Search Engine Result Pages (SERPs). Draft paper. HTML
  9. ^ Yu Suzuki, Masatoshi Yoshikawa: Assessing quality score of Wikipedia article using mutual evaluation of editors and texts. Proceedings of the 22nd ACM international Conference on information & knowledge management, Pages 1727–1732. ACM New York, NY, USA 2013. http://dx.doi.org/10.1145/2505515.2505610 Closed access icon
  10. ^ Raíza Hanada, Marco Cristo, Maria da Graça Campos Pimentel: How do metrics of link analysis correlate to quality, relevance and popularity in Wikipedia? Proceedings of the 19th Brazilian symposium on Multimedia and the web, Pages 105–112. ACM New York, NY, USA 2013 http://dx.doi.org/10.1145/2526188.2526198 Closed access icon
  11. ^ Marcelo Yuji Himoro, Raíza Hanada, Marco Cristo, Maria da Graça Campos Pimentel: An investigation of the relationship between the amount of extra-textual data and the quality of Wikipedia articles. Proceedings of the 19th Brazilian symposium on Multimedia and the web, Pages 333–336. ACM New York, NY, USA 2013 http://dx.doi.org/10.1145/2526188.2526218 Closed access icon
  12. ^ Sook Lim: Does Formal Authority Still Matter in the Age of Wisdom of Crowds?: Perceived Credibility, Peer and Professor Endorsement in Relation to College Students’ Wikipedia Use for Academic Purposes. ASIST 2013, November 1–6, 2013, Montreal, Quebec, Canada. PDF
  13. ^ Jinyoung Kim: Wikipedians From Mars: Female Students’ Perceptions Toward Wikipedia. ASIST 2013, November 1–6, 2013, Montreal, Quebec, Canada. PDF
  14. ^ Eckert, S.; Steiner, L. (2013). "(Re)triggering Backlash: Responses to News About Wikipedia's Gender Gap". Journal of Communication Inquiry. 37 (4): 284. doi:10.1177/0196859913505618. Closed access icon
  15. ^ Khoi-Nguyen Tran, Peter Christen: Identifying multilingual Wikipedia articles based on cross language similarity and activity. Proceedings of the 22nd ACM international conference on information & knowledge management, Pages 1485–1488, ACM New York, NY, USA 2013, http://dx.doi.org/10.1145/2505515.2507825 Closed access icon
  16. ^ Thamar Solorio, Ragib Hasan, Mainul Mizan: Sockpuppet Detection in Wikipedia: A Corpus of Real-World Deceptive Writing for Linking Identities http://arxiv.org/abs/1310.6772
Supplementary references:
  1. ^ Seward, Zachary M. (13 August 2009). "Here's the AP Document We've Been Writing About". Nieman Lab, Harvard University. ...the new routine of Twitter-to-Google-to-Wikipedia contrasts sharply with the behavior of users in August of 1997...
  2. ^ "Page view statistics for Wikimedia projects". Wikimedia Foundation. Retrieved 25 November 2013.

















Wikipedia:Wikipedia Signpost/2013-12-04/Recent_research