Skip to content
  1. When Michelle Obama said “When they go low, we go high” during the 2016 Democratic National Convention, the rallying cry was met with thunderous applause. But according to a new analysis by School of Journalism graduate student Floris Wu and lecturer Aleszu Bajak, negativity proved to be a winning Democratic strategy on social media in the lead-up to the 2018 midterm elections.

In a recent analysis published in Roll Call, Bajak and Wu analyzed the sentiment of Twitter messages from Democrat and Republican senatorial candidates on the campaign trail in 2018. The analysis was conducted by looking at the Twitter timelines of 68 Senatorial candidates and scoring their average positivity. The sentiment of these tweets was also calculated by running them through a sentiment dictionary developed at the Computational Story Lab at the University of Vermont and validated using a machine learning model that classifies tweets as either positive or negative.

Take a look at the scatterplot of the sentiment used by candidates in 2018 below:

Take a look at the full analysis at: