Introduction
Engagement with reporting mechanisms (‘flagging’) online is an increasingly important part of monitoring and addressing harmful content at scale. However, users may not flag content routinely enough, and when they do engage they may be highly biased by group identity and political beliefs. Across several highly-powered and pre-registered online experiments, we examine the extent of social bias in people’s flagging of hate speech and abuse. Preliminary results show that while people do reliably engage with reporting mechanisms, they are significantly more likely to do so when hate and abuse targets in-group compared to out-group members.
To try and improve user reporting of online hate and abuse, in this series of experiments we also provide the first empirical tests of novel interventions aimed to improve flagging behaviours. The interventions centre on enhancing user knowledge about when and how to flag content, and enhancing transparency about what happens to content that is flagged.
Project aims
The aims of the project are to understand the extent of social bias present when people flag content online, and to provide an evidence base for interventions that may enhance user engagement with flagging mechanisms, both in quantity and quality of reporting. This will improve user engagement with reporting mechanisms online, an important part of addressing harmful content at scale.
Organisers
