Research:Teahouse group dynamics

From Meta, a Wikimedia project coordination wiki
23:00, 25 July 2016 (UTC)
Duration:  2016-May – ??

This page documents a research project in progress.
Information may be incomplete and change as the project progresses.
Please contact the project lead before formally citing or reusing results from this page.

Who socializes the socializers?

The Teahouse is a project for socializing new editors in the ways of the English Wikipedia community. But the Teahouse itself is a community of practice with its own local norms and values, some of which are differ from those of Wikipedia as a whole. What strategies does the Teahouse use to socialize new hosts, and to how successful are these strategies at maintaining its founding norms?



This project will consist of an examination of the interactions among hosts, and between hosts and guests, over the 4+ year history of the Teahouse. The goal of the project is to understand how the Teahouse has maintained itself as a group—with common goals and a distinct group identity—in the face of internal and external changes. What we learn about the Teahouse can help us understand how other groups within the Wikimedia Movement—from individual WikiProjects to whole wiki communities—adapt to environmental changes, and inform the design of tools and policies to support those groups.


The Teahouse continues to be an active WikiProject four years after it was created,[1] and has been shown to increase new editor retention. Many experienced Wikipedians participate as Teahouse hosts. Some have been participating for years. However, there has been a great deal of turnover since February 2011, when the Teahouse was launched. Most of the original host cohort are no longer active participants, and many hosts from subsequent cohorts have also come and gone.

The founding hosts helped define a set of norms and procedures for running the Teahouse within the first ~6 months of the project. Resources created by the Teahouse founders, such as the host expectations, are still available for hosts who joined later to use learn how to participate. However, new hosts weren't part of the process of creating those norms—do they understand them the same way as the founders, and abide by them to the same extent?

How does a new host become socialized in the ways of the Teahouse? There is some evidence of explicit socialization on the Teahouse talkpages. For example, new hosts asking questions about how to respond to certain guests, or experienced hosts reminding people to use @mention in their replies to guests. However, as of mid-2016, there is a lot less of this kind of explicit articulation work and norm enforcement happening on the talkpages than there used to be. And yet, on the surface hosts' behavior on the Teahouse Q&A board still shows a great degree of alignment with the original set of norms. Most questioners are welcomed. They generally receive a @ping to let them know that someone has responded to them. And they seem to receive detailed and personalized replies (not just a list of links), even if the question they asked is one that others have asked many times before.

Social norms and behavior[edit]

We examine the degree to which two of the Host Expectations ("welcome everyone" and "avoid over-linking") are influenced by injunctive and descriptive norms. Injunctive norms, like the Teahouse host expectations, describe what people ought to do. Descriptive norms reflect observed behavior—what people actually do.[2] Group dynamics generally reflect a mix of injunctive and descriptive norms. Whether hosts perceive a particular behavior (like welcoming everyone who asks a question) as governed primarily by injunctive or descriptive norms has consequences. If a new host starts welcoming questioners because they believe that this is the required/accepted thing for a host to do (injunctive norm), they will likely continue to do so even if they see other hosts fail to include a welcome in their responses. On the other hand, if new hosts only welcome because they see other people doing so (descriptive norm), they are less likely to continue to perform this behavior if others stop doing it.

If welcoming behavior is viewed by new hosts primarily as an injunctive norm, it is likely to persist over time. If it's perceived as a descriptive norm, it's more likely to degrade over time: if existing hosts fall out of the habit, newer hosts will be less likely to adopt it, potentially creating a positive feedback loop that results in the norm becoming strictly optional or even disappearing from the Teahouse completely.

Likewise, if hosts who join the Teahouse have already internalized a different set of norms around the best way to answer a question, the Teahouse-specific norms for answering questions will be less salient to them, and they will be less likely to abide by them even if they have read the Host Expectations and see other people following these norms. Thus, an influx of new hosts from other Wikipedia help forums, like the Help Desk, could dilute the descriptive norms around welcoming and policy linking on the Teahouse and lead to an evolution away from these norms over time.

Research questions[edit]

  1. Do hosts who have been exposed to the Host Expectations via the host profile creation process include a welcome in their reply more often, and do they include fewer links in their replies? (Injunctive norms)
  2. When new hosts join, do they tend to follow the example of recent replies by other hosts in their own welcoming and linking behavior? (descriptive norms)
  3. Do hosts who have been exposed to the Host Expectations and who join during periods where other hosts are welcoming more and linking less tend to adopt these behaviors more than those for whom only one of the above conditions holds true? (congruent injunctive and descriptive norms)
  4. Do hosts with prior experience answering questions at the Wikipedia Help Desk, which has different norms around welcoming and linking behavior, welcome less and link more than hosts who have not participated in that forum? (conflicting injunctive and descriptive norms)


We will analyze the text of questions and answers from the Teahouse archive, as well as characteristics of people who participate in the Teahouse: hosts' registration date, the number of questions they answered, whether or not they've created a profile, whether or not they have answered questions at the Help Desk.

Policy, Ethics and Human Subjects Research[edit]

All text and metadata related to participation in the Teahouse are drawn from publicly-available datasets using API and/or database queries. Interviews with Teahouse editors are conducted according to the Wikimedia Foundation's Privacy policy and Data retention guidelines. The purpose of the study will be disclosed to interview participants. The primary investigator is a founding member of the Teahouse with a history of good-faith volunteer contribution to the project.


Once your study completes, describe the results an their implications here. Don't forget to make status=complete above when you are done.

See also[edit]

Subpages of this page[edit]

Pages with the prefix 'Teahouse group dynamics' in the 'Research' and 'Research talk' namespaces:

Research talk:


  1. Morgan, Jonathan T.; Bouterse, Siko; Walls, Heather; Stierch, Sarah (2013). "Tea and Sympathy: Crafting Positive New User Experiences on Wikipedia". Proceedings of the 2013 Conference on Computer Supported Cooperative Work. CSCW '13 (New York, NY, USA: ACM): 839–848. ISBN 9781450313315. doi:10.1145/2441776.2441871. 
  2. Kallgren, Carl A.; Reno, Raymond R.; Cialdini, Robert B. (2000-10-01). "A Focus Theory of Normative Conduct: When Norms Do and Do not Affect Behavior". Personality and Social Psychology Bulletin 26 (8): 1002–1012. ISSN 0146-1672. doi:10.1177/01461672002610009.