Research Data_Exploring Social Media Information Seeking During the COVID-19 Pandemic to Inform Health and Science Communication_Sage Open.csv

This study utilized a non-experimental descriptive survey design to examine social media information-seeking behaviors during the COVID-19 outbreak, particularly during lockdown periods. The objectives were to describe perceptions of COVID-19 information on social media, explore the platforms used during lockdown, identify groups of connections on social media, and determine if platform use varied based on connected groups. To gather data on information-seeking behaviors, an online survey was administered via Qualtrics, reaching 1,048 respondents in the United States through non-probability opt-in sampling. The survey included the perceptions of the information availability scale (Woolley & Propst, 2005) and an information-seeking behavior scale (Timmers & Glas, 2010) the information availability scale (Woolley & Propst, 2005), and some researcher-adapted Likert-type scales. The results revealed that more than 70% of respondents felt overwhelmed while searching for COVID-19 information, encountered difficulties in accessing and interpreting additional information, and sometimes even avoided news about the pandemic. Among social media platforms, Facebook, Instagram, and Twitter were the most popular for obtaining COVID-19 information. Notably, Facebook emerged as the most widely used platform during lockdowns. Furthermore, respondents primarily utilized Facebook to connect with friends and family during the pandemic, and those with larger social networks tended to access social media platforms more frequently. These findings highlight the significant role of Facebook in disseminating reliable information during the COVID-19 pandemic. They also emphasize the importance of implementing strategies to help individuals navigate the overwhelming amount of information, including misinformation, present on social media platforms, particularly during times of crisis. It is worth noting that there is limited generalizability due to US-centric sample.