As the 2020 election draws closer, political content on social media is becoming more and more prevalent. A group of Penn State researchers is examining spontaneous sharing of political content on Facebook.
According to principal investigator S. Shyam Sundar, James P. Jimirro Professor of Media Effects at Penn State, the rushed nature of social media use may lead users to be less deliberate and more spontaneous in their interactions. “This is resulting in the widespread phenomenon of sharing without clicking. When online users see a link or headline that appears to be aligned with their ideology, they are more likely to share it rather than scrutinizing the contents of the link or story.”
The researchers want to assess if users are being more honest in expressing their stance when they share without clicking political content on Facebook. They will look at the types of content and other characteristics of posts, such as number of likes, that are more often shared without clicking, and how closely aligned the content is with the user’s political stance.
The research will further our understanding of the role played by communication technologies and online political discussion via social media, and may also be able to inform interventions to foster more thoughtful content sharing.
The project is being funded by a Social Media and Democracy Grant (awarded by the Social Science Research Council (SSRC)), and includes access to data provided by Facebook about user behaviors. The SSRC funding arrangement with Facebook and Social Science One will also allow researchers to see the websites users have shared, unique characteristics of the URLs, and some demographic information about the users, such as their age, gender, location, and political leanings.
“We expect to see extreme political content to be shared without clicking, which would be indicative of spontaneity,” Sundar explained. “We also plan to pair this data with concurrent investigations on the role played by device, and expect mobile use to be associated with more spontaneous sharing of political content, given its small screen and the rushed nature of its use.”
Facebook is the largest social media platform in history, and social scientists who study how democracies function have long wanted access to demographic data on the platforms’ two billion users, said Sundar. However, privacy concerns have held up Facebook data from being released to researchers for study. Facebook has been working on implementing a statistical technique called differential privacy to ensure that user information is scrambled by adding noise to the datasets provided to the researchers.
“For a social scientist used to dealing with data from a few hundred cases, the prospect of analyzing a dataset with 17 trillion cell values relating to the sharing of nearly 38 million URLs by 2 billion Facebook users is certainly exciting and worth all the effort to ensure privacy,” said Sundar.
Penn State’s Computational and Spatial Analysis Core (CSA), part of the Social Science Research Institute and the Population Research Institute, will be analyzing the data for the project. Guangqing Chi, director of the CSA and professor of rural sociology, demography, and public health sciences in the College of Agricultural Sciences, says that social media data provide significant opportunities to study social problems that cannot be easily addressed by traditional data.
“Although we’ve seen a rapid increase in the use of social media data for social science research, we know little about how users behave. This timely project will contribute important knowledge about political content sharing on social media. Our CSA team is excited to collaborate with Dr. Sundar’s researchers in processing and analyzing the 17 trillion cell values.”
Other researchers on the project include Junjun Yin, assistant research professor, Social Science Research Institute and Population Research Institute, and associate in the CSA; and graduate students Jinping Wang, Eugene Cho and Meng Qi Liao in the Donald P. Bellisario College of Communications.