Blinded by the Spite? Path Model of Political Attitudes, Selectivity, and Social Media

ABSTRACT

Despite fears that selective exposure and selective avoidance could deepen polarization and negatively affect the democratic process, few studies have directly studied this phenomenon. This study explores whether selective exposure and avoidance to blogs, social network sites, and Twitter directly influence confidence in Congress and the president or more indirectly through polarization. This study suggests that fears of selective exposure, selective avoidance, and polarization infecting the democratic process appear overstated. First, polarization was positively related to confidence in Congress and the president. Second, selective exposure to social media sites strengthens confidence in the president and in Congress. Twitter boosts confidence in Congress. Third, selective avoidance had a negative influence on other measures, which suggests people seek both information that challenges their views as well as ones that supports them. Finally, selective exposure and avoidance proved weak indicators of polarization. Instead, strength of partisanship is the stronger predictor of confidence in Congress and the president.

KEYWORDS:

• Selective exposure
• selective avoidance
• polarization
• blogs
• Twitter
• social network sites

Notes

¹ Announcements were also placed on a researcher-created Facebook group called “Facebook Users and Political Polarization” and on existing groups of a variety of political philosophies. The authors posted status updates announcing the survey with a request to share the update with others. One of the authors also posted on a note on Facebook about the survey and tagged friends, urging them to complete it. Survey announcements were also posted on academia.edu, LinkedIn, and the University of Texas LinkedIn group. Respondents were solicited on Twitter and asked to retweet the survey URL. Respondents were contacted through searches such as “Twitter and Politics”; Facebook and Politics”; and hashtags on leading Republican, Democrat and Libertarian political figures.

² Cooperation Rate: As defined by American Association for Public Opinion Research “Standard Definitions”: “The proportion of all cases interviewed of all eligible units ever contacted.” The survey was housed on MRInterview, a web-based survey site. MRInterview tracks the number of people who clicked on the survey link from various sources and the number who actually completed the survey. MRInterview calculated a cooperation rate of 27.8% for this survey. A meta-analysis of online survey response rates found that our response rate was typical, with some studies falling below 10% (Manfreda, Bosnjak, Berzelak, Haas, & Vehovar, 2008).