FollowBias

Abstract

Networked gatekeepers on social media increasingly influence which people and groups receive media attention. Many unknowingly direct much greater attention to men than to women. Can technologies support these gatekeepers to follow their own values of equality? Theories of value consistency suggest that confronting people with inconsistencies between values and behavior can prompt behavior change. In this paper, we introduce FollowBias, a novel system that offers feedback on the percentage of women that users follow on Twitter. We conduct field deployments of FollowBias with 61 and 78 participants, exploring differences between their values and behavior, their explanations of those differences, and their changes in behavior. In the first, FollowBias users had a 45 percentage point greater chance of increasing the percentage of women followed over one week. In the second, we fail to find an effect. We also offer findings on political and ethical trade-offs in designing systems for behavior change toward equality.