Disemvoweling

CAPTCHA

Began as a way of reading books, but bots soon excelled.

Defensiveness

behavior which occurs when an individual perceives threat or anticipates threat in the group. (Gibb 1961, p. 141)

spends energy defending themselves rather than focusing on the common task; often reciprocated

… defensive recipients distort what they receive. As [he becomes] more defensive, he becomes less and less able to perceive accurately the motives, the values, and the emotions of the sender. (Gibb 1961, p. 142)

QIC go-around

ex. Web scavenger hunt

1. Form three groups. Within each group, pairs should address a portion of the claims.
2. Find an example of a design criteria from either:
   1. http://reddit.com/
   2. http://en.wikipedia.org (in Wikipedia: namespace) (remember search tips)
3. Document in respective Google Docs.
4. You have 15 minutes

Limiting Effects of Bad Behavior

1. Moderation systems that pre-screen, degrade, label, move, or remove inappropriate messages limit the damage they cause.
2. Redirecting inappropriate posts to other places will create less resistance than removing them.
3. Consistently applied moderation criteria, a chance to argue once case, and appeal procedures increase the legitimacy and thus the effectiveness of moderation decisions.

4. Moderation decided by people who are members of the community, are impartial, and have limited or rotating power will be perceived as more legitimate and thus be more effective.
5. Reversion tools limit the damage disrupters can inflict in production communities.
6. Filters or influence limits can limit the damage of shill raters in recommender systems, but only at the cost of ignoring some useful information from honest raters.
7. A widely followed norm of ignoring trolls will limit the damage they can do.

Coerced Compliance

8. Activity quotas allow people to participate in a community, but prevent repetitive, spam-like activity.
9. Gags and bans can limit the continuing damage of a bad actor, but only if it is hard for the bad actor to use a different account or if the ban is disguised.
10. Consistently applied criteria for gags and bans, a chance to argue one’s case
11. Paying to take actions in the community with currency accumulated through normal participation will reduce the ability for trolls and manipulators to act.
12. Limiting fake accounts with CAPTCHAs or identity checks limits automated attacks.

Making Norms Clear and Salient

13. Publicly displaying examples of appropriate behavior on the site will show members what is expected and increase their adherence to those expectations.
14. Publicly contrasting examples of inappropriate behavior in the context of a descriptive norm of appropriate behavior will highlight the descriptive norm and increase people’s adherence to it.
15. Publicly displaying many examples of inappropriate behavior on the site will lead members to believe this is common and expected.

16. Displaying feedback of members to others increases members’ knowledge of community norms and compliance with them; formal feedback is more effective than informal feedback.
17. In large communities, displaying statistics that highlight the prevalence of normative behavior will increase members’ adherence to normative behavior.
18. Explicit rules and guidelines increase the ability for community members to know the norms, especially when it is less clear what others think is acceptable.
19. Prominently displayed guidelines may convey a descriptive norm that the guidelines are not always followed.
20. Offering people reminders at the point of an action that may violate norms will reduce the number of offenses.

Questions?

• What are 3 important Kraut & Resnick design tactics for limiting bad behavior?
• Which of Gibb’s six traits is most important online?
• What norms do you think is most important at Wikipedia?

▶