Examining the Impact of COVID-19 on Food Desert and Food Swamp Twitter Discourse: An Exploration of Engagement-Driven Selective Sampling

One out of every 10 homes experienced food insecurity at some point in 2021, defined as inadequate access to food. The COVID-19 pandemic globally increased the risk of food insecurity, and Twitter provides a timely way for people to express their personal viewpoints. This study paired the medium of social media online discourse on Twitter with the topic of food insecurity by searching for social media posts that mention the two locational reporting categories of food insecurity: food deserts, defined as poor access to affordable, healthy produce, and food swamps, defined similarly, with emphasized access to cheap, fast-food choices. This study provides additional understanding of the impact of COVID-19 on food insecurity and the conversations that surround the topic, as well as how to utilize big data analysis to analyze Twitter discourse on the platform more efficiently. The purpose of this study was to better understand food desert and food swamp Twitter discourse through big data analysis and explore the impact opinion leaders and the COVID-19 pandemic had on online public Twitter conversations. Using total population census data and quantitative content analysis, I also sought to see the impact of selective sampling on the census data set. Results found few studies on agricultural education and communication-based research studies implementing Twitter API for food insecurity discourse, and none containing census data sets. The COVID-19 timeline was found to have an impact on the engagement of food desert and food swamp tweets, as well as the inclusion of combative language. The inclusion of hashtags, media, user verification, and combative language all had additional impacts to tweet engagement. Engagement-driven selective sampling was found to impact the frequency of tweet characteristics in the data sets. More research is needed to support the findings of this study, and the full census data sets are available in the appendices of this study to support further research on food swamp and food desert discourse online. Additional research is needed on the justification and communication theory behind engagement-driven selective sampling.