Community Atomization on Semi-Private Online Platforms

Project Lead / Student Investigator: Erika Melder
Principal Investigator: Dr. Michael Ann DeVito
Co-Investigator: Dr. Ada Lerner

Background

We define a semi-private online platform as a platform where users adopt a persistent identity to connect with other users in multiple disconnected spaces. One example is the text and voice chat application Discord, which is one of the most popular social media platforms, reporting over 150 million monthly active users as of March 2024. Discord allows its users to create servers, which are independently operated collections of text and voice chat channels. Users may then join others’ servers to socialize.

There are two factors in this arrangement which invite further exploration. First, users have a single shared account that is used to interact with all servers. Second, server owners can control user permissions on their server: they can vet who is able to join, promote or demote users to have elevated privileges (such as the ability to moderate other users), and temporarily or permanently remove accounts from their servers. The policies around permissions vary from server to server, and are based on the risk model employed by the server staff.

These factors in combination introduce tradeoffs for users looking to find spaces where they fit in. Notably, at-risk communities are forced to have permeable boundaries to allow in marginalized users while blocking those who would break the community guidelines. Therefore, instances hosting marginalized users face tension between allowing unfettered access to their server, and running a restricted server which prioritizes protecting the existing community. This contributes to the phenomenon of atomization, where servers offering community and support become too isolated to discover or too difficult to establish oneself in, preventing users from finding them.

Project Description

This project seeks to determine how the structures of semi-private online platforms, such as Discord, affect community boundaries. In particular, we are interested in how governance and access decisions are impacted, from the perspective of both regular users and staff members. We are also interested in whether these structures contribute to atomization of communities on the Discord platform. We will conduct an interview study of Discord users to determine how membership and governance decisions are made and enforced, and how this in turn affects community-building and community-finding on the platform.

Research Questions

  • How do membership and leadership practices positively or negatively impact community-building and community-finding on Discord?
  • How are membership and governance decisions on Discord informed by users’ privacy and security needs?
  • How does the technical structure of Discord affect the procedures and outcomes of these user actions?

Expected Outcomes

Our goal with this project is to develop user-informed guidance for Discord users and administrators.

How to Participate

Recruitment for this study is now closed. The inclusion criteria are listed here for documentation purposes. In order to participate, you must be:

  • 18+ years old
  • An active Discord user
  • Able to conduct an interview in English
  • Located in the US, Canada, the UK, or the European Economic Area
    • Individuals outside these regions may still be able to conduct an interview with us. Please let the researchers know and we will do our best to work with you.

Here is the consent form for the study.

If you have any questions about the study, you can contact Erika Melder via one of the following channels: