Twitter’s Reaction to Mask Mandates: A Network Analysis

Our world is made up of networks. Everything from the neurological pathways that are fundamental to brain functions to the power grid keeping the world moving are examples of how networks are essential to society. Perhaps most impressive of all, they can store an extensive amount of information in an organized manner. This information can be analyzed to understand relationships between different entities, gain predictive probabilities, and identify possible areas of community. Throughout the scope of our paper, we dive into one of the biggest networks online by analyzing information directly from the social media platform of Twitter. We explore tweets from several different states regarding the mask mandates that resulted from the 2020 pandemic and examine the before and after impact it had on different users. Additionally, we divided up these tweets using a number of keywords and hashtags that were associated with anti-mask and pro-mask connotation. We compared these pro and anti mask networks with a sentiment analysis and therefore quantified the “sentiment” with calculated scores. We then looked at detailed summary statistics of our different networks to gain insight on each state’s pro and anti mask tweet networks. We continued our analysis by applying community detection algorithms for possible subgraphs within our larger network and comparing them to each other. Finally, we utilized our findings to identify some of the most connected users within our networks and analyze them from the user level. In this comprehensive analysis of a Twitter network, we explore a wide range of different network diagnostics from both the tweet and the user perspective to search for potential relationships.