New project to examine social media's impact on political attitudes and behavior

New York University Faculty of Arts and Science professors Richard Bonneau, John Jost, Jonathan Nagler, and Joshua Tucker have established a project to examine the impact of social media use on political attitudes and participation by applying methods from a range of academic disciplines.

"Social media use has been linked to the spread of political protests throughout the globe, but, unlike during demonstrations from previous eras, today's communication tools, such as Twitter and Facebook, give us real-time accounts of protests documented by the participants themselves," explained Tucker.

The project, Social Media and Political Participation (SMaPP), is funded by a three-year, $1 million grant from the National Science Foundation and will be composed of faculty from NYU's Wilf Family Department of Politics, Department of Psychology, Department of Biology, and Department of Russian and Slavic Studies as well as NYU's Courant Institute of Mathematical Sciences.

Tucker added that SMaPP can also help inform a growing debate on social media's impact on political participation.

"Despite a preponderance of anecdotes about the effects of social media, the reality is that we know little about whether social media actually facilitates political activity, and, if so, how it does so," he said.

While social media's influence on political participation and attitudes remains in question, Twitter, Facebook, and Instagram undoubtedly offer amounts of data that far exceed earlier research methods. Whereas behavioral scientists from approximately 1940 to 2000 relied on surveys of hundreds or perhaps thousands of respondents, social media can provide researchers with datasets involving tens of millions of participants.

"The ubiquity of social media offers astonishing opportunities to gauge the determinants of collective action that were not previously possible—and may call into question traditional assumptions about the nature of human communication and persuasion," noted Jost.

The project will rely on both survey data and publicly available social media data such as "Tweets" to address a range of questions concerning the causal processes that shape political participation.

"Because many users access social media repeatedly, they create a rich time-series dataset of individual-level behavior," explained Bonneau. "To extract the potential of these data, we will adapt computer learning models, such as those developed in the field of genomics, and integrate these with more established theories and methods in the behavioral sciences."

For example, on May 1, 2012, a day on which Occupy Wall Street protests occurred in New York City, the research team collected more than 80,000 tweets featuring #OWS and related hashtags.

"The ability to harvest massive quantities of real-world transmissions and to connect these transmissions to meaningful behavioral outcomes will almost surely transform our theoretical and empirical understanding of social influence and political behavior," explained Nagler.