News packaging during a pandemic: A computational analysis of news diffusion via Facebook

Abstract

Facebook remains the most important platform where social media editors package and try to ‘sell’ media outlets’ online news articles to audiences. In one of the first studies of its kind, we assess how this practice was effectuated during the first year of the COVID-19 pandemic. We use computational analysis to determine the polarity, subjectivity and use of some linguistics features in the status messages of 140,359 Facebook posts of 17 mainstream and alternative news titles from Flanders (Belgium) between March 2020 and 2021. Among other things, we find that status messages score considerably higher than headlines in terms of polarity and subjectivity, and that they, along with the use of question and interrogation marks, peaked in the first months of the pandemic. We contextualise our findings within existing scholarship and wider trends in increasingly digitised and globalised media societies.