User:EpochFail/On-wiki research

From Meta, a Wikimedia project coordination wiki

Scholarly research of Wikipedia is useful for understanding the encyclopedia's content, readers, editors, history, current state, and future. These results also yield important knowledge applicable to other open content communities. In addition to driving scholarly knowledge of such systems, this work can also give results that can improve Wikimedia projects themselves. Much valuable research cannot be done without allowing researchers to interact with Wikimedia communities. This guideline exists to describe how researchers should behave when performing activities that affect Wikimedia projects or their members.

How can research contribute?[edit]

For a more complete explanation of researchers activities, see: en:User:EpochFail/What_are_these_researchers_doing_in_my_Wikipedia

Expanding articles is not the only way to contribute to Wikimedia projects. There is room for many different types of contribution: copy edits, content edits, template construction, vandal fighting, bot management, etc. Through the use of the scientific method, researchers contribute to Wikimedia projects by extended understanding about the community and bringing state-of-the-art technologies to editors.

What do researchers need to know[edit]

  • Respect the work of the community. Above all, research activities must respect the work of a community. This means that research should not disrupt editor's work and must respect the policies and norms of the community. Breaching experiments are not acceptable without wide community consensus and such a consensus for a breaching activity is very unlikely.
  • Seek approval. Research should seek feedback and approval from the community they intend to study before they perform any research activities. Community members who are likely to be affected by research activities should be canvassed and consensus about the activities should be found. For research subject recruitment, see User:EpochFail/Research_recruitment.
  • Operate transparently. Researchers should operate transparently. Make it apparent to those editors you interact with when operating under the role of researcher. Publishing your research protocol is recommended so that editors will know what to expect when they interact with you. Some types of experiment will require that details be withheld from the research participants, but you'll need to disclose enough for affected editors to know what you are planning and what that might imply for them and the community.
  • Make your results a contribution. (Something about wikimedia projects and openness). Members of Wikimedia communities will be more accepting of your activities if they can benefit from results of your work. If copyright does not allow you to distribute the results of your work freely, consider publishing a summary of your results that community members can access.

What do editors need to know[edit]

  • Researchers contribute by doing research. The intention of scholarly research is to build and share knowledge about the world. This goal is in line with the mission of the Wikimedia Foundation and it's projects. With help from the community in which they interact, researchers can occupy a valuable role.
  • Researchers need your help. Many researchers are not Wikimedians and might need help understanding the right way to interact with a community. Traditionally, the academic community functions very differently from open communities, so the policies and culture of wikis can be unfamiliar. Assume good faith. Researchers who end up disrupting your work most likely did not mean to and will stop if you ask. Try to help them find a way to conduct their activities that does not cause a disruption.
  • Participate. Some research requires participation like taking a survey or testing a new interface modification.

Participant Recruitment[edit]

Recruiting research participants is essential for performing many types of valuable research. Surveys (eg. the yearly editor survey) and interviews (eg. "Becoming Wikipedian" by Bryant et al.) require that a sample of interested editors be questioned. Understanding the effects of a change to the system (eg. Usability initiative, SuggestBot, WikiTrust, WikiDashboard, NICE, HAPPI, etc) requires that a sample of users be recruited to test the change. In both of these cases, a representative sample of users is essential for attaining a valid results. Although, in some cases a sampling editors by posting on a discussion board or mailing list is acceptable, in many cases no central communication system exists for recruiting from a desired pool of users.

Recruiting a representative sample of users without a central communication system can be difficult. Policies against unsolicited personal communication (WP:SPAM, WP:CANVASS, etc) that are intended to protect editors from advertisement spam and other types of disruptive communication can make it impossible to recruit a random sample of users while respecting policies.

The (Subject Recruitment Discussion Group|SRAG|Whatever) was formed to operate as a venue to allow editors to gain consensus for their recruitment activities from the communities from which they recruit and allow for random, unsolicited messages to be sent if the research activities are deemed acceptable by the community.