Harassment & Hate Speech Against the 2018 Migrant Caravan

Our Investigations Lab students researched hate speech targeting the 2018 migrant caravan made up of thousands of people from Central America. Students scraped tweets associated with the caravan’s trending hashtags on Twitter, with the purpose of analyzing them to answer the following research questions:

1. To what degree was the conversation on the caravan and immigration on Twitter powered by artificial intelligence and/or machine learning bot accounts?
2. How did the U.S. President’s tweets influence the behavior of highly centralized bot and non-bot accounts?
3. What is the prevalence and impact of highly centralized bot and non-bot Twitter accounts in spreading disinformation, harassment, and hate speech on the caravan and immigration conversation on Twitter?
4. What is the prevalence and impact of highly centralized bot and non-bot Twitter accounts in spreading and perpetuating political and ethnic divisiveness on the caravan and immigration conversation on Twitter?

The Lab team was tasked with coding approximately 3,000 tweets in order to flag categories including false/misleading images and videos taken out of context; political divisiveness; ethnic divisiveness; and instigation of violence.