Research:Online Community Conduct Policies/Quora

Quora is a question-and-answer website where questions are asked, answered, edited and organized by its community of users. The firm was founded in June 2009, and the website was made available to the public on June 21, 2010. Quora aggregates questions and answers to topics. Users can collaborate by editing questions and suggesting edits to other users' answers. It also allows individual contributors to see statistics on views, upvotes, and shares of questions and answers. Quora is ranked 133 globally with about 17M unique page views a month.

Community characteristics

Quora had an estimated 500,000 registered users ("Quorans"), as of January 2011; more recent statistics on community size are difficult to come by, as Quora does not release them. The site's visitors come primarily from India (35.5%) and the United States (29.4%), and are significantly more likely than elsewhere on the internet to be male (and, correspondingly, significantly less likely to be female). Quora users are also significantly more likely to have their highest level of education be graduate school.

In addition to typical site metrics available in places like Alexa, the site's registration method (suggesting users associate a social media account and requiring their real name on their Quora account) makes it possible for enterprising members of the Quora community to do complex research regarding qualities like gender (see, for instance, the body of work by Laura Hale on that topic here) based on the social media accounts linked to Quora accounts.

Conduct policy

See also: Quora terms of service, Quora policies and guidelines

Quora's main user behavior expectations are listed in their "Be nice, be respectful" policy. This policy states that users should treat each other with "civility, respect, and consideration." It breaks unacceptable behavior down into sub-categories including:

Attacking people or content

People should be civil and respectful in disagreement with others and should not:

1. make attacks or otherwise disparage other people,
2. refer to other people's content with insulting or disrespectful language, or
3. harass others on the site. These behaviors hurt the Quora community and deter users from helping to create a better resource.

The policy also notes that Quora holds its users to a higher behavioral standard than some other social media sites, as rather than connecting friends who already know each other, Quora's Q-and-A threads tend to connect unrelated strangers who simply share an interest in a topic. The "Be nice, be respectful" policy provides a set of examples of unacceptable types of personal attacks.

Harassment

Quora defines harassment as behavior that is intended to have a negative effect on, or make the site environment unwelcoming to, another user. Its behavior policy provides a non-exclusive list of behavior that may be interpreted as harassment:

1. making threats,
2. repeated annoying and unwanted contacts,
3. posting the personal information of another person,
4. using sexually explicit, profane, adversarial, or flirtatious language toward another person if such language would likely make the person uncomfortable.

Hate speech and racial slurs

The policy prohibits content or "tone" that attacks people based on qualities such as "race, gender, religion, nationality, ethnicity, political group, sexual orientation or another similar characteristic." It also provides a link to Wikipedia's List of ethnic slurs, noting that any word included in that list is prohibited on Quora except in the case of asking a "sincere" question about the word's etymology.

Site policy is also to remove from the site the accounts and submitted content of any user who states themselves to be, or can be proven to be, a member of any organization listed on the United States State Department list of Foreign Terrorist Organizations. This includes organizations such as al-Qa’ida and Boko Haram.

Suicide

Quora automatically adds a banner with "here's how to get help" resources to any question on the site that is tagged under the topic "suicide". In addition, users who notice a comment or question that indicates a threat of suicide can manually add the "suicide" topic to force the banner. Users can also contact Quora staff directly using the "Contact us" form.

Retaliation

The policy explicitly states that violating the "be nice, be respectful" policy is not ok even in cases where one is reacting to provocation or to someone else having broken the policy first.

History of the policy

Quora's early conduct and moderation policies drew heavily from those of Wikipedia and other FOSS communities like Python. In the time since, both the conduct and moderation policies have been adapted to better suit Quora's and its community's unique needs.

The site keeps an edit history for all questions asked on it, including those related to official policy. This link, for instance, provides a changelog of the site's "What are the main policies and guidelines" page between 2012 and the current time; here one can see the development of the "Be nice, be respectful" policy from one short paragraph in 2011 to today's highly detailed and interlinked version of the policy. Today, the site notes that "[p]olicy ideas originate from individual Quora users and employees—via questions on the site, private messages, and email. That feedback is consolidated and reviewed by Quora Moderation staff, who determine which, if any, policies need to be created or improved. These proposed policies are also reviewed with trusted users on the site before being posted  publicly. As a result of this process, Quora’s policies are evolving—they are not set in stone."

The site's moderation has evolved in tandem with its policies, moving from relying largely on unpaid, trusted volunteers to process problematic or flagged content circa 2011, to relying solely on a paid moderation team employed by Quora to perform these functions circa 2015. As described in a pair of 2011 blog posts by Marc Bodnick (part 1, part 2), Quora based some of its early moderation strategies on Wikipedia's administrator policies, acknowledging that social media sites like Quora and Wikipedia need to rely on a trusted class of users who are granted additional permissions in order to keep the site under control.

Quora implemented this "trusted user" moderation strategy with two levels of permissions for trusted volunteers: reviewers, who had the right to collapse and flag potentially problematic questions and answers, and admins, who had the additional rights to edit and delete questions and answers, and suspend and ban users from the site. Both groups were appointed by Quora, the company, rather than by the community. The additional privileges assigned to Quora administrators came with requirements for transparency and accountability for those users, obligating them to log rationales for their deletion and blocking actions; in addition, moderation decisions were generally not made on one administrator's judgment but rather on the consensus of the group of active administrators.

By 2013, the administrator selection process had been streamlined into appointment of a new cohort of active administrators and de-adminship of inactive administrators every six months. In January 2015, however, the site rewrote its moderation process from the ground up, merging the former "administrator" and "reviewer" groups into a single "trusted reporter" group. The erstwhile admins and reviewers in this group no longer have the ability to delete or collapse content or block users; rather, they now have the single ability to submit fast-tracked reports about content or users who violate site policy. These reports are then reviewed and acted upon by the paid Quora Moderation team, who retain the ability to delete and block.

Policy enforcement

As with many social media sites, Quora prefers its users to be their own first-line defense against problematic behavior, allowing them to block other users from interacting with them. A user who is blocked cannot follow, receive notifications about, ping, send requests or invitations to, or up- or down-vote the person who blocked them. This action also automatically deletes any previous interactions or notification settings between blocker and blockee.

If blocking the user is insufficient, or if the behavior cannot be stopped through a user-to-user block, users are encouraged to report harassment and other problematic behavior to the Quora moderation team. Third-party reporting (reports made by users other than the victimized user) is encouraged, and Quora maintains a group of Trusted Reporters, unpaid site users whose reports have a history of accuracy and are therefore fast-tracked through moderation queues. Circa 2014, the site also claimed to be developing a human-enhanced AI system called Content Review to automate some types of moderation. Though the system's details are not public, it involves some amount of bot editing of questions and comments that meet certain criteria, as well as some amount of human review.

Both moderation staff and the Content Review bot have the capability to edit users' questions. Moderation staff can also delete questions and answers and issue time-limited "edit-blocks" to user accounts, the latter preventing the suspended user from editing the site but not from reading it. In extreme cases, users may be banned entirely from the site by moderation staff, losing access to their account. Moderation staff does not speak publicly about the reason behind a given account's block or ban.

Moderator actions are generally appealed privately by emailing a dedicated "appeals" address (or by doing the same through the site's "Contact us" form). Users are asked to supply the basis of the action they are appealing, the content that led to the original moderation decision, and any other important details. The site is also trialing a new appeals system, currently in use only for appeals about answer collapses (a common moderation action that minimizes but does not delete problematic content). In this new workflow, users are asked to (re-)read the policy they violated and to state that either they have edited the problematic content of their collapsed answer, or that they believe the moderation decision was incorrect and that their content did not violate the cited policy.

Analysis of policy: strengths and weaknesses

External links

• Quora Moderation blog