The final ECREA 2022 session for today starts with Soyeon Jin, whose focus is on the European immigration debate. She notes that Europeans’ attitudes towards immigration have improved over the years, yet there also seems to be an increasing amount of controversial debate around; what is going on here? Traditional assumptions are that more negative sentiment produces more negative messages on social media, but this may not be the only dynamic here; people with more extreme views may simply have become more active in their social media posting, too.
To further explore this, Soyeon’s work focusses on reaction buttons on Facebook, grouping Love and Haha in one vector, Wow into another, and Sad and Angry into a third. Such reactions may indicate agreement or disagreement – and their distribution might indicate polarisation (where there are two distinct peaks in responses) or simple diversity of views (a more equal distribution).
Agreement should also lead to less follow-on discussion (there is less need for further debate), while a lack of agreement should produce more discussion in an attempt to resolve those differences or at least win the argument. Yet polarisation might also produce Spiral of Silence effects, as people are less willing to continue to argue if they feel their position is unwelcome.
Soyeon managed to get access to a URL sharing dataset from Social Science One (!), and from this identified immigration-related URLs, but these needed to further cleaned to remove content related to animal migration, server migration, and other irrelevant topics. In the end this left some 23,000 URLs, and the aggregate reactions to these URLs were largely diverse. This also had a significant impact on further likes and shares, while evidence of polarisation had a significant influence on further likes, shares, and comments.