Learning from 2016: predicting elections through Twitter
What happened during the news coverage for the 2016 election cycle(s) is a case study that will be studied for years to come. The failure to accurately predict Brexit or Donald Trump’s success in the US presidential election left everybody perplexed as neither, especially the latter, was ever predicted to occur. This research investigates the failures of the media to predict Trump’s election and investigates whether Twitter can be used in the place of existing procedures to correctly predict elections. Data obtained through this research shows that online discussion of Trump by Twitter users was higher than discussion of Hillary Clinton during four critical events leading up to election day. This paper argues that this data can be used to make successful predictions for election outcomes. The findings and conclusion help underscore the importance of this type of research, not only in terms of predicting elections via Twitter, but also in terms of how such insights might benefit the media and its audiences.