Elsevier

Computers in Human Behavior

Volume 52, November 2015, Pages 211-218
Computers in Human Behavior

The role of news in promoting political disagreement on social media

https://doi.org/10.1016/j.chb.2015.06.011Get rights and content

Highlights

  • Network size and frequency of use are positively related to news use.

  • News use is positively related to political messaging and monitoring.

  • Messaging and monitoring are positively related to political disagreement.

  • Messaging and monitoring mediate the relationship between news use and disagreement.

Abstract

Determining whether and how social media networks expose individuals to political disagreement is critical to understanding how individuals experience civil society in the digital age. Recent research from the United States and Europe shows that, all else equal, heavy social media users are exposed to more political disagreement on social media than light users. The present study seeks to elaborate on these findings in the context of Colombia. In doing so, it adds depth to existing theory about social media and political disagreement by outlining a process for how social media use results in exposure to disagreement and the role played by news. Results from path analysis show that (a) news use on social media acts as a link between general use and disagreement and (b) political engagement mediates the relationship between news use and disagreement. Results are discussed in light of existing literature and possibilities for further research.

Introduction

As social media proliferate on a global scale, often with substantial political impact (e.g., Eltantawy and Weist, 2011, Harlow, 2012, Howard and Parks, 2012, Lim, 2012, Valenzuela et al., 2012), important questions arise about their role in processes related to political expression and the public sphere. Recent arguments assert that the modern individual is situated within an egocentric public sphere (Papacharissi, 2009, Rojas, 2014, Wojcieszak and Rojas, 2011). Individuals are increasingly embedded in loose, personalized networks of affiliation and exchange (Benkler, 2006, Rainie and Wellman, 2012), and they are connected to these networks through digital media that promote expression and interactivity (Bennett, 2008, boyd et al., 2007, Loader and Mercea, 2011). These digital communication networks link the private spaces of individuals with the public sphere, that is, they connect people with civil society (Friedland et al., 2006, Loader and Mercea, 2011, Rojas, 2014). Determining whether and how political disagreement occurs within these networks is therefore critical to understanding how individuals experience civil society in the digital age (Barbera, 2014, Brundidge, 2010, Kim, 2011, Kim et al., 2013).

Recent research from the United States and Europe shows evidence that social media use is positively related to political disagreement on social media (Barbera, 2014, Kim, 2011, Kim et al., 2013, Mitchell et al., 2014). All else equal, heavy social media users are exposed to more political disagreement than light users. The present study seeks to elaborate on these findings in the context of Colombia. In doing so, it adds depth to existing theory about social media and political disagreement by outlining a process for how social media use results in exposure to political disagreement. This model rests on two broad claims. First, social media use facilitates exposure to news posted by a wider array of individuals and organizations. Second, political disagreement occurs when individuals engage with social opinion about news.

Existing theory about why social media use results in political disagreement rests on two observations: (a) Social media afford opportunities to share information and express personal opinions and (b) social media diversify communication within egocentric networks through the articulation of weak tie relationships (Barbera, 2014, Brundidge, 2010, Kim et al., 2013). In simpler terms, social media expose people to more information from more sources than they would otherwise be exposed to. Selectivity does little to counteract these forces. Social selectivity is multidimensional and not limited to political choice (Kim et al., 2013). Informational selectivity, meanwhile, is more likely to be politically motivated. However, people do not necessarily avoid crosscutting news media online (Garrett, 2009) and interpersonal recommendations on social media often trump partisan media cues (Messing & Westwood, 2014). Thus, social media might inadvertently expose individuals to political disagreement (Brundidge, 2010).

Importantly, this theory is built on the concept of communicative diversity rather than social network diversity. Social media may not diversify social networks, but they do diversify communication that occurs within them. This is a subtle but important distinction that avoids the pitfalls of counterarguments based on social norms of connectivity. Rather than use social media to meet new people, most people use them to articulate existing social connections (boyd et al., 2007, Ellison et al., 2007), although it must be said that some social media sites (e.g., Twitter) are more conducive to network expansion than others (e.g., Facebook or Instagram). But even while the primary role of social media is to articulate existing social networks rather than to expand them, articulation still diversifies communication in comparison to interpersonal contexts, which are limited by geographic space (Brundidge, 2010, Huckfeldt et al., 2004), and anonymous online contexts, in which relatively homogeneous political communication occurs (Hill and Hughes, 1998, Wojcieszak, 2008).

When it comes to political communication on social media, news is (one of) the primary source of public information, and recent literature emphasizes the role of news in promoting communicative diversity on social media. For example, Lee, Choi, Kim, and Kim (2014) show that social media diversify communication networks, in part, through news use (see also, Barbera, 2014), which implies, of course, that social media promote news use. This conclusion is generally borne out by observational analysis: Research shows a positive relationship between social media use and news use on social media in various political contexts (Gil de Zúñiga et al., 2012, Valenzuela et al., 2012).

Part of the explanation for the relationship between general use and news use has to do with network size and structure. Larger, more diffuse networks are better at spreading information in social networks because they contain more weak ties (Granovetter, 1973). Research shows a consistently positive association between network size and content diffusion (Adar and Adamic, 2005, Bakshy et al., 2009, Cha et al., 2009). Meanwhile, a large-scale Facebook experiment shows that the number of friends posting a link to a story increased the probability of sharing that story (Bakshy, Rosenn, Marlow, & Adamic, 2012). Finally, social media network size is positively related to relevant behaviors, including political participation and/or group formation, commonly thought to result from information exposure (e.g., Backstrom et al., 2006, Gil de Zúñiga et al., 2012, Valenzuela et al., 2012). And while other structural characteristics of networks are also important for information diffusion (Centola, 2010), network size facilitates the influence of many of these structures (Horowitz & Malkhi, 2003).

Information-sharing affordances of social media represent another explanation for the relationship between use and news use (Loader & Mercea, 2011). Posting news is relatively uncommon among the average user, but it is very common among politically involved users (Glynn, Huge, & Hoffman, 2012). In fact, approximately 20–30% of social media users, who some have called “power users,” account for substantially more content than typical users (Hampton, Goulet, Marlow, & Raine, 2014a). In other words, a few users post a lot of news, which means that the average user is exposed to more news than they post. Once again, this conclusion is borne out in research: About 50% of U.S. adult web users get news from social media, which is approximately the same proportion as those who watch TV news (Barbera, 2014, Mitchell et al., 2014). In Colombia, the percentages are even higher, where, according to this study, 66% of social media users report using either Facebook or Twitter for news.

Given the above literature, there is good reason to believe that network size and frequency of use will be positively related to news use on social media. However, it is also important to remember that the way individuals use media matters when it comes to its effects (Shah et al., 2001a, Shah et al., 2001b). The Internet generally contributes to the fragmentation of the news audience based on interests or preferences (Prior, 2005). The politically disinterested might use social media less, particularly those social media oriented toward news (e.g., Twitter). They might also use social media differently, choosing not to read or pay attention to news. Therefore, it is important to consider the role of political interest when examining the relationship between general social media use and news use.

News use promotes engagement with politics and public affairs. For example, news use is associated with political learning (Eveland, Shah, & Kwak, 2003), political discussion (Shah, Cho, Eveland, & Kwak, 2005), cognitive reflection (Cho et al., 2009), and political participation (Shah et al., 2001). These studies, which belong the family of models knows as the communication mediation models, suggest that media effects are largely indirect and mediated through indicators of cognitive and/or discursive engagement with information. Political talk, or informal discussion about politics in everyday life, figures centrally into many of these models (e.g., Shah et al., 2005, Nah and Shah, 2006) and is considered to be a key facilitator of civic and/or political participation.

News use on social media also promotes political expression (Gil de Zúñiga et al., 2012, Valenzuela et al., 2012). In fact, social media afford new forms of political messaging (Gil de Zúñiga et al., 2012) built around the virtual spaces that news stories provide online. And even while recent research shows that commenting on news articles is relatively rare on social media (Hampton, Raine, Lu, Dwyer, Shin, & Purcell, 2014b), when it does occur, it can have influential effects on information processing and its subsequent outcomes, including political participation (Park, 2013, Yamamoto et al., 2013, Zhang et al., 2010) and exposure to disagreement (Kim, 2011, Kim et al., 2013).

But political talk–or other forms of political messaging–is not the only way to engage with news and public affairs information on social media. People have always used news to monitor social opinion about public issues (Noelle-Neumann, 1984[1993]), and social media visualize social information about others in extended egocentric networks (see, e.g., Walther, Van der Heide, Kim, Westerman, & Tong, 2008). Arguably, social media users have more information with which to monitor their social networks for political opinions as compared to other communicative settings (Ho and McLeod, 2008, Schulz and Roessler, 2012). Moreover, social media juxtapose information from mass-mediated and interpersonal sources (Walther et al., 2011). These messages could interact to influence information evaluation and perceptions of others’ opinions (Anderson et al., 2013, Paek et al., 2013, Walther et al., 2010, Xu, 2013).

This literature implies that, on social media, interaction with others via political talk or political messaging is not necessary in order to engage with the news and public affairs. Individuals can get a sense of the conversations around stories without participating in those conversations themselves. Therefore, it is important to consider how political monitoring on social media–that is, learning about friend’s or follower’s political opinions via social media content–acts as an additional form of engagement that is interrelated with political messaging.

The now-classic studies on political disagreement defined it either as the lack of agreement between two discussants (e.g., Huckfeldt et al., 2004) or the perception of disagreement resulting from interaction (e.g., Mutz, 2006). These conceptual differences are important because differences in operationalization can yield different results (Klofstad, Sokhey, & McClurg, 2012). Specifically, comparative measures typically find more evidence of disagreement, while perceptual measures typically find less. These discrepancies are also reflected in studies of social media, with some employing perceptual measures of exposure to crosscutting views (e.g., Kim, 2011, Kim et al., 2013) and others using comparative measures of social media ties (e.g., Barbera, 2014). This study opts for the perceptual conceptualization of disagreement. As Mutz (2006) argues, it is the experience of disagreement that matters most when it comes to political behavior. While the lack of agreement has a subtle influence (Huckfeldt et al., 2004), disagreement has deeper and more lasting effects (Mutz, 2006).

Another advantage of the perceptual approach is that it helps to distinguish disagreement from a related antecedent, communicative diversity. Disagreement is more likely in diverse communication networks (Barbera, 2014), but engagement with diverse information makes it more likely that individuals will perceive a discrepancy between themselves and others (Mutz, 2006). Hypothetically, it therefore makes sense to predict that disagreement on social media will be an outcome of engagement with others about public affairs information. Remembering the indirect nature of media effects predicted by the communication mediation models (e.g., Shah et al., 2005), social media news use should have an indirect relationship with political disagreement on social media, such that it is mediated through political talk and political monitoring.

The theoretical model in Fig. 1 illustrates the expectations outlined above. Based on the discussion of news use on social media, the model predicts that social media network size (γ7) and social media use (γ3) will be positively related to social media news use, and that these relationships will exist above and beyond the influence of political interest (γ5). Based on the discussion of engagement with news on social media, the model predicts that social media news use will be positively related to both political talk (β1) and political monitoring (β2) on social media. In order to isolate these relationships, paths were included between the antecedent variables and political talk/monitoring. Finally, and based on the discussion of disagreement as the outcome of engagement with news on social media, the model predicts that social media news use will have direct (β3) and indirect positive relationships with political disagreement on social media through political talk (β5) and political monitoring (β6), which will themselves be interrelated (β4).

Section snippets

Context of study

This study uses the most current data from an ongoing, biennial survey of Colombian adults in urban areas. Colombia’s history of violence and polarizing politics has resulted in relatively low levels of trust in the government and democratic processes, and the rise of efforts to promote transparency and responsiveness in local government, particularly in urban areas (Rojas, 2010). In Colombia, citizens are beginning to peacefully reengage in political processes in the aftermath of more than 50 

Results

Before testing the full model, OLS regression was used to assess the relationships between social media political disagreement and the other social media variables of interest. Table 1 shows results for a series of models in additive blocks that estimated changes in explained variance (all results reported in this section were statistically significant with p < .05, unless otherwise noted). The first model, which included none of the social media predictors, explained about 13% of the variance in

Discussion

This study made several predictions based on the model illustrated in Fig. 1. First, social media use and network size would be positively related to news use on social media. Second, social media news use would be positively related to political messaging and monitoring on social media. Third, political messaging and monitoring would mediate the relationship between news use and political disagreement on social media. Finally, these relationships would exist independently of political

References (62)

  • Y. Benkler

    The wealth of networks: how social production transforms markets and freedom

    (2006)
  • W.L. Bennett

    Changing citizenship in the digital age

  • D.M. Boyd et al.

    Social network sites: definition, history, and scholarship

    Journal of Computer-Mediated Communication

    (2007)
  • J. Brundidge

    Encountering ‘‘difference” in the contemporary public sphere: the contribution of the Internet to the heterogeneity of political discussion networks

    Journal of Communication

    (2010)
  • Cha, M., Mislove, A., & Gummadi, K.P. (2009). A measurement-driven analysis of information propagation in the flickr...
  • J. Cho et al.

    Campaigns, reflection, and deliberation: Advancing an O-S-R-O-R model of communication effects

    Communication Theory

    (2009)
  • D. Centola

    The spread of behavior in an online social network experiment

    Science

    (2010)
  • N.B. Ellison et al.

    The benefits of Facebook “friends:” social capital and college students’ use of online social network sites

    Journal of Computer-Mediated Communication

    (2007)
  • Eltantawy, N. & Weist, J.B. (2011). Social media in the Egyptian revolution: Reconsidering resource mobilization...
  • W.P. Eveland et al.

    Assessing causality in the cognitive mediation model: a panel study of motivations, information processing, ad learning during Campaign 2000

    Communication Research

    (2003)
  • L. Friedland et al.

    The networked public sphere

    Javnost – The Public

    (2006)
  • R.K. Garrett

    Politically motivated reinforcement seeking: reframing the selective exposure debate

    Journal of Communication

    (2009)
  • H. Gil de Zúñiga et al.

    Social media use for news and individuals’ social capital, civic engagement, and political participation

    Journal of Computer-Mediated Communication

    (2012)
  • Granovetter, M.S. (1973). The strength of weak ties. American Journal of Sociology, 78(6), 1360–1380...
  • Hampton, K., Goulet, L. S., Marlow, C. & Raine, L. (2014). Why most Facebook users get more than they give. Report for...
  • Hampton, K., Raine, L., Lu, W., Dwyer, M., Shin, I., & Purcell, K. (2014). Social media and the “spiral of silence”....
  • S. Harlow

    Social media and social movements: Facebook and an online Guatemalan justice movement that moved offline

    New Media and Society

    (2012)
  • A.F. Hayes

    Introduction to mediation, moderation, and conditional process analysis: a regression-based approach

    (2013)
  • K.A. Hill et al.

    Cyberpolitics: citizen activism in the age of the internet (people, passions, and power)

    (1998)
  • S.S. Ho et al.

    Social-psychological influences on opinion expression in face-to-face and computer-mediated communication

    Communication Research

    (2008)
  • P.N. Howard et al.

    Social media and political change: capacity, constraint, and consequence

    Journal of Communication

    (2012)
  • Cited by (0)

    Survey was approved by UW’s Social and Behavioral Sciences IRB and funded by the University of Wisconsin-Madison Graduate School.

    View full text