Volume 20 Numbers 3 & 4, 2010
Voice to the people: Media users’ perspectives on selective exposure and avoidance
Magdalena E. Wojcieszak
Selective exposure is among the most extensively researched concepts in communications, although it was not until relatively recently that scholars demonstrated that media users are not passive audience but active participants in media attendance, reception, and interpretation. Asking “What do people do with the media” rather than “What do the media do to people” (Katz, 1959, p. 2) has challenged the powerful media effects paradigm, bringing users’ decisions to scholarly attention (Katz, 2001). This question is especially useful in the contemporary media environment which not only facilitates but requires active participation.
Although the very notion of selectivity is intrinsically linked to audience decisions, individual motives behind selecting or avoiding certain content have been insufficiently researched within the scholarship on selective exposure. Also, although the literature recognizes that interests and ideology motivate media use and mediate the effects, scholars have primarily focused on the choices per se rather than on what occurs before or after. Moreover, although some studies have differentiated between selective exposure and avoidance, showing that partisans seek consonant content but might not avoid dissimilar views (e.g., Garrett, 2009), the processes behind these phenomena have not been closely scrutinized. Questions such as: Why do people in general, and strong partisans in particular, use the news media as they do? How do people perceive what they read, hear and see? Do they avoid oppositional outlets and why? and How do user s react to challenging opinions? are underdeveloped in the literature on selectivity. This study begins to address this gap by asking students active in on-campus partisan organizations about their experiences with news media.
Research on Selective Exposure and Avoidance
Katz’s “Mass Communications Research and the Study of Popular Culture” (1959) made a strong argument for the “active” audience, calling for attention to what people do with the media rather than how the media affect individuals. This shift in research focus has led scholars to note that people are not uniformly affected by the totalizing media, but are rather agents who select sources and interpret content. Consequently, the audience role has been conceptualized as encompassing a “seeker; consultant; browser; respondent; interlocutor; or conversationalist” (McQuail, 1997, p. 129). Underlying all these roles is selectivity.
Selectivity is defined as “a process of nonrandom selection of media-related alternatives” (Levy, 1987, p. 268), with the “nonrandomness” resulting from individual attitudes and expectations about the messages’ utility. That is, content selection and evaluation are influenced by interests or political convictions (Ball-Rokeach, Rokeach, & Grube, 1984). With regard to selection stage, individuals attempt to increase their exposure to messages and sources that are consistent with their attitudes. They are more likely to read, listen to or watch information that confirms their beliefs and less likely to attend to information that challenges their position (Lazarsfeld, Berelson, & Gaudet, 1944). People also tend to select outlets that support their views over less supportive alternatives, a tendency occurring when choice is limited and increasing with an increase in media options (Lazarsfeld, 1948; Stroud, 2008).
Some research suggests that selectivity may be especially pronounced among partisans. Partisans prefer magazines, newspapers, television programs and talk shows that support their views (Mutz & Martin, 2001), and obtain news from consonant online sources (Best, Chmielewski, & Krueger, 2005). This preference for supporting views increases with heightened commitment to individual position (Jonas et al., 2001; Taber & Lodge, 2006). Thus, strong partisans tend to visit their favored candidate’s websites (Garrett, 2009) and select consonant outlets for such apolitical topics as travel and leisure and especially for such contentious issues as the Iraq war (Iyengar & Hahn, 2009).
Although some scholars see preference for supportive content and aversion to dissimilar views as two attributes of a single phenomenon (Festinger, 1957), others distinguish between selective exposure, or seeking out confirmatory opinions, and selective avoidance, or actively filtering out opposing viewpoints. Often, challenging information is useful for argumentation or planning (Frey, 1986; Garrett, 2009). In a similar vein, dissimilar sources might be welcome when they address relevant issues, indicating that what might be driving content selection is not partisan selectivity related to psychological defenses but rather issue selectivity guided by interests and pragmatism (Iyengar , Hahn, and Prior, 2001). Furthermore, whether or not people avoid challenging content might depend on individual factors. Specifically, characteristics that predict attit ude-consistent attention, such as knowledge or curiosity, also predict counter-attitudinal exposure. Also, political involvement increases news use, no matter the source’s position (Chaffee et al., 2001). Attitude importance, political interest, and partisanship strength are also related to selecting counter-attitudinal messages (Knobloch-Westerwick & Meng, 2009), indicating that people with these characteristics may be sufficiently confident in their views so that to seek out counterarguments (Albarracín & Mitchell, 2004).
Yet other research explains media use with de facto selectivity whereby social circumstances expose people to consonant information. Exposure to opinion-reinforcing views would thus result from decisions unrelated to ideology, such as interests, personal networks or social context. According to this account high income voters encounter more pro-Republican messages because their friends and neighbors are primarily Republican, not because they avoid Democratic messages (Sears & Freedman, 1967). All in all, research conceptualizes selectivity as involving motivated exposure to content in order to obtain reinforcement, regards selectivity as content selection related to interest or utility, and sees selectivity as relatively passive and driven by reasons unrelated to partisanship.
This scholarship, however, has primary focused on media choice, rather than on the motivations that lead to the choice or the interpretations that follow. Research on uses and gratifications can shed some light on these issues because it has scrutinized precisely the needs and interests that motivate media use, aiming to discover why “people use media” and what “they use them for” (McQuail, 1997, p. 72). This research has provided insight about individual motivations for media use, such as escapism or surveillance, and has also categorized these motivations into diversion, social utility, personal identity or cultural norm reinforcement, among others (Katz, Blumler & Gurevitch, 1974; McQuail, Blumler, & Brown, 1972). Uses and gratifications scholarship, however, has not spoken to partisan selective exposure and avoidance because it has mostly analyzed media in general (newspapers, cable television, computer games, and so forth) or certain genres (such as soap operas, news programs, or entertainment shows) rather than politically or ideologically consonant or dissonant media content (e.g. Bryant & Zillman, 1984; Lemish, 1985; Perse & Dunn, 1998; Rubin, 1984). Nevertheless, it is worthwhile to explore whether the broad typologies put forth by uses and gratifications can offer a general framework from which to approach phenomena related to partisan selective exposure and selective avoidance.
Also, selectivity research has provided important insight into individual media choice. What it has missed, however, are the individuals themselves. That is, people have not been directly asked about their motivations behind, experiences with, and interpretations of news media content in the selectivity context. Although people are said to prefer consonant information, they have not been asked about our reflections on media selection and reception. Media users are also told that they avoid opposing outlets because they dislike having their views challenged, but those users have not been asked whether this is the reason and how it feels to have their beliefs contradicted. It is also alleged that people turn to oppositional sources if they are open-minded and motivated to learn, but people have not directly described how they experience such exposure.
Although quantitative methods have elucidated many relationships between individual position and media selection, surveys or experiments are largely inapplicable to addressing the issues raised above. This is because “[t]he shift in accent from external observable action to internal action processes … and the related need for more insight into these processes places certain demands on the research strategy” (Renckstorf, 1996, p. 30). Qualitative methods are better suited to meet those demands; open-ended, in-depth, retrospective interviews are particularly appropriate for capturing media use dynamics (LaBennett, 2006). Although not free from limitations, interviews credit the audience with self awareness and thus allow exploring the expectations, motivations, and meanings that people assign to media (see Ginsburg, Abu-Lughod, & Larkin, 2002; Hermans & van Snippenburg, 1996; Spitulnik, 2002). In other words, because communication activities are about “people’s perceptions and observations about their lived-in worlds ... it stands to reason that it would be important to ask them to express their experiences and perceptions” (Stromer-Galley, 2002).
This is precisely what this study aims to do. It brings people back into the scholarship on selectivity, relying on in-depth interviews with young partisan activists to address the following questions: To what sources do informants turn for political information? What are the motivations behind and gratifications obtained from selecting these sources? Do informants expose themselves primarily to ideologically consonant news outlets? Do they avoid outlets that present counter-attitudinal information? If so, why? How do they react to messages that challenge their political beliefs? To address these questions and analyze the rich textual data, this study uses the grounded theory approach (Glaser & Strauss, 1967), and in the process attends to the ways in which uses and gratifications may help explain some questions about partisan selectivity.
This analysis draws on in-depth, loosely structured interviews with 12 undergraduate students active in major on-campus Democratic and Republican organizations at a large Ivy League university (see Appendix A). Those individuals were chosen as informants because they are politically involved and committed to liberal or conservative ideology, and as such could be expected to select consonant sources. On the other hand, their activism and issue commitment could encourage dissonant media use, speaking to the processes underlying selective avoidance. Due to their organizational engagement, informants could also provide insight into the role that interests play in selectivity. Thanks to their education, moreover, they could be expected to elaborate on their media activities. Finally, those students, active in partisan organizations within an Ivy League university, working for presidential campaigns, and frequently from prominent families, will one da y become the political elite and it is worthwhile to explore their media use patterns. Although the sample is not free from limitations, which are addressed later, these interviews can shed light on the important processes analyzed here.
The interviews took place on or around campus in November and December 2006. Because Midterm Elections in the U.S. fell within this time period, heightened party loyalties could exacerbate selectivity. The interviews were long, open-ended and loosely structured. This approach made it possible to ask follow-up questions, focus on issue elaboration, deviate from a question in order to probe an interesting point, and let the informants provide abundant and sometimes seemingly unrelated information (Wengraf, 2001). Interviews revolved around news source selection and the reasons behind it, opinions on the content, perceived information utility, exposure to outlets seen as oppositional, and reactions to dissonant information (see Appendix B). The interviews ranged from 65 to 120 minutes in length, and added up to 22 hours of recording.
Using the grounded theory approach (Glaser & Strauss, 1967), textual data were carefully examined and the accounts provided by each informant were then fragmented into pertinent clusters that speak to the research foci and to the questions asked during the interviews. These textual fragments were read, and then compared and contrasted in order to identify patterns and inconsistencies within each individual’s narrative. The fragments from each category were also assembled across all the informants. Thus, responses from similar individuals could be weaved together in order to identify recurring patterns and similarities as well as to spot idiosyncrasies and inconsistencies between the informants and their accounts. These concurrent processes, juxtaposing textual fragments within a single narrative and across various narratives, made it possible to reassemble informants’ words into an explanation that accounts for complex patterns, including idiosyncrasies and contradictions. In the process, the data were reduced, by focusing on the information relevant to the topic and discarding extraneous information. Also, analytical distance accompanied the process to identify hidden meanings or contradictory motivations that, when assembled, portray media use and partisan selectivity in a nuanced and comprehensive way (Charmaz, 2006).
Data limitations. Before presenting the results, two limitations should be acknowledged (both are addressed more fully in the conclusion). First, due to limited access to cognitive and emotional states and/or due to purposeful attempts to mislead the interviewer, the informants may not reliably report their motivations behind media use and their reactions to certain sources. As a result, some findings may tell us more about how the informants think and explain their experiences than about the actual experiences. Secondly, the sample is not only small, but also quite homogeneous with regard to age, education, race, ethnicity, and social status (9 informants were male, 9 white, all in their early 20s). This substantially limits the conclusions that can be drawn from these conversations. While these limitations need to be kept in mind, the insight provided by exploring partisan selectivity via in-depth interviews – especially given the dearth o f such research in this context – justifies the effort.
Profitable Soundbites - Opinions on the News Media Environment
Before exploring the patterns underlying selectivity, motivations, and interpretations, it is worthwhile to present the informants’ general behavior and opinions on news media environment. Likely due to their upbringing and political involvement, the informants keep up with the news. For the most part, they turn to major conglomerates, such as CNN, Fox, The New York Times, or The Washington Street Journal. Although these sources are mentioned most frequently, all informants perceive the U.S. media system in a negative light. Democrats and Republicans alike criticize the media for focusing on profit, relying on soundbites, and lacking in-depth reporting.
Echoing some communication theorists (McChesney, 2004; Patterson, 1993), the informants believe that the news media aim to increase revenues, rather than contribute to knowledgeable and effective citizenry. “It is a business that they have to make entertaining” or “They will put on whatever the people want to hear ... otherwise they will go out of business,” informants say. Those political sophisticates also criticize news media for reporting soundbites rather than substantively covering “real issues.” Informants complain that the media prioritize empty rhetoric over debates on concrete solutions. Some go as far as blaming the media’s reliance on soundbites for problems in the system: “Everything that is wrong with the American politics can in some way be traced back to the media,” because running for an office “becomes a marketing campaign. Candidates ar e a product that needs to be marketed through the media.”
Informants further say that mainstream sources only offer basic information, and hence use them solely to obtain “the major four or five events of the day” or “run through the headlines to find out what the issues are.” Beyond this surveillance function, mainstream news media are seen as useless because they lack specialized coverage and substantiated information that those activists need. They say: “If you want to just keep up to date, that’s fine, you can watch CNN for half an hour. But you are not going to really understand,” or “You can get all the soundbites and the headlines. But you are not going to get any in-depth analysis.” For in depth analyses, the informants turn to specialized outlets and the internet, which is seen as most convenient and which is often contrasted with the inefficient traditional media. They regard going online as easier than picking up a newspaper, especially “with limited time a nd working on the computer anyway.” Informants also dislike “flipping through channels,” and are unwilling to “set my schedule around the TV news especially with the 24 hour news cycle.” The internet also facilitates locating specific issues not covered in the mainstream media. Informants thus read blogs about venture capital or foreign policy, Israeli news sites, and services that compile articles from diverse newspapers.
Does the fragmented media environment, with its source abundance and content diversity, invite partisan selectivity? According to the informants, the answer is yes. They generally turn to consonant alternative outlets that are explicit about their political position. Conservatives go to Drudge Report (http://www.drudgereport.com/), which features news and commentary, GrassrootsPA (http://grassrootspa.com/), a blog run by the Young Conservatives of Pennsylvania, and Truth Caucus (http://www.truthcaucus.com/), a blog by the Federation of College Republicans. They also listen to conservative talk radio hosts, such as Rush Limbaugh, Sean Hannity, or Michael Smerconish. All Republicans mention at least one of these sources. Democrats, in contrast, use explicitly liberal sources less frequently. Only some read self-categorized liberal magazines, e.g., The Atlantic and The Nation, or blogs, such as Daily Kos (http://www.dailykos.com/)and MyDD (http://mydd.com/). Only one interviewee tuned to the liberal talk radio network, Air America.
Online blogs, talk radio, and other alternative outlets, which make their political stances explicit, seem to facilitate partisan selectivity partly due to the a priori affinity between the source and the audience. Their overt political leanings also make the motivated and purposive selective exposure easily detectable. But what about self-selection to attitude-consistent mainstream sources? A noteworthy caveat emerging from the interviews pertains to the problems inherent in categorizing such sources. Should researchers or media users decide whether a Democrat watching CNN is exercising selectivity or whether Fox exploits conservative biases? Do Republican Fox viewers or Democratic NPR listeners make Fox conservative and NPR liberal? Or perhaps all mainstream sources fall into the center? A 20-year old Democrat, Joe (the names have been changed), offers his own take on this issue:
In other words, determining whether selectivity occurs in the mainstream news media context requires asking whether mainstream sources espouse a certain ideological agenda. The answer, in turn, depends on the criteria used to assess this agenda. There are at least three approaches. First, we could accept the audience’s categorizations. In so doing, we would conclude that partisans do not turn to consonant mainstream sources but rather use impartial media. That is, because individual leanings influence the perception of an outlet’s leanings, ideologues do not see their preferred sources as biased (Vallone, Ross, & Lepper, 1985). Simply put, as one interviewee notes, “People think that agreeing with their opinion is the same as being neutral.” Republicans thus claim that Fox is balanced because about “half of their staff is Democrat” or note that it is seen a s conservative but say: “it is more about the issues” and “it doesn’t necessarily give you a more conservative perspective.” Similarly one Democrat argues that The New York Times criticizing George W. Bush does not indicate liberalism; rather the lack of criticism “would be biased. So that means that they are presenting a balanced story.” Another Democrat sees CNN as “neither liberal nor biased” and yet another argues that NPR, as a public broadcaster, is “as good as it gets.” Following this approach, only the alternative sources would allow for purposive partisan selectivity.
Perhaps we should claim that if a Democrat sees his/her sources as neutral but a Republican perceives them as liberal, these sources are biased? If, conversely, a Republican claims to have an impartial media diet, which a Democrat considers conservative, the Republican in fact attends to conservative sources? If we rely on such assessments, we will conclude that – across the board – partisans from the party opposed to the judge’s own are selective. Because partisans on both sides of the spectrum tend to perceive media content as “hostile” to their own position (Vallone et al., 1985), the Republicans say that CNN’s anchors “are overtly liberal and do not hide it,” and Democrats claim that Fox presents a “spun Republican conservative agenda.” This is well represented by Jack, a 21-year old Republican business student:
Perhaps we should turn to moderate partisans to determine political leanings of mainstream outlets and, based on these assessments, determine what constitutes partisan selectivity in the mainstream media? The moderates often note that their preferred sources have some bias; and the way those informants categorize mainstream media resembles some academic categorizations (e.g., Stroud, 2008). Anna, a 22-year-old Republican senior who favors civil unions, notices that the Fox’s Iraq War coverage favors the Republicans inasmuch as reporters “don’t ask ‘why are we there to begin with’” Sam, a 21-year old Republican who grew up in a religious family and went to a Catholic high school concludes that Fox might be partial as it has “guys like O’Reilly, Neil Cavuto, Fred Barnes, Bill Kristol, they are very conservative.” Kristy, a 20-year old Democrat, says “I don’t feel t hat The New York Times has a liberal bias in their general news articles but in their opinion pieces most definitely,” and a 21-year old social liberal Chris states: “I tend to place CNN as objective, may be very very slightly to the left.” Generally, those Democrats who see CNN, NPR, or The New York Times as leaning to the left quickly note that these leanings are not as overt as Fox’s conservative bias.
All in all, the interviews do not provide an unequivocal answer to the question “are partisans selectively using mainstream media?,” and are noteworthy for this very reason. While informants acknowledge self-selection to consonant alternative sources, which are explicit about their political stances, the problems inherent in categorizing mainstream news do not allow us to conclude that partisan selectivity is exercised in this context. At best, the interviews suggest that – with regard to the mainstream sources – partisan selectivity does not seem to be a consciously recognized activity, and that turning to certain sources appears to be motivated by basic surveillance function.
Motivations for Selective Exposure
With regard to the motivations, the interviews suggest that selective exposure may be driven by familiarity with a source, its entertainment value, and a desire to learn, rather than by a need to validate own views and obtain psychological comfort. Nevertheless, these non-political motivations may be related to partisanship and ideology. These various nuanced interconnections, elucidated by studies on socialization, biased processing, and hostile media effect, are explicated in the discussion section.
First, selectivity results from habits and familiarity with certain sources. All but three informants belong to the same party their parents belong to and attend to the same sources that were used at home. Mike, a 21-year old Republican, says that his father, a campaign contributor, would listen to Rush Limbaugh and watch O’Railly Factor. In addition to conservative blogs, these sources are among Mike’s favorites. Sam, who likes Neal Cavuto, started watching Fox with his parents. Similarly Anna regularly visits foxnews.com and also recalls watching Hannity and Colmes with her mother. Sean, a 19-year-old Democrat reads The New York Times and The Economist, to which his family subscribed. Katie, a 22-year old Republican who watched news during family dinners, underscores these patterns saying: “I kind of absorbed my parents’ values ... although this was mostly an unspoken absorp tion.”
Secondly, the informants see their preferred sources as entertaining. For example Mike, a Republican who attributes the unpopularity of the Vietnam war to the limited media choice in the 1970s, listens to conservative radio hosts, such as Sean Hannity, Rush Limbaugh, or Michael Smerconish because they are enjoyable and unpretentious. Mike sees them as “original,” “down to earth,” and “human in addition to the political.” He says: “I have an impression that this is just a normal person carrying out a conversation and presenting news at the same time. They don’t have the East Coast mentality” (by which he means an intellectual or elitist tone). Also Sam watches Fox and listens to conservative radio hosts. He likes these sources because they are amusing and “because there is certain attitude surrounding the news that’s unique.”
Attitude-consistent exposure is also motivated by a desire to learn. This rationale comes close to wanting to confirm one’s own views, inasmuch as the informants learn about their preferred worldview and gain arguments to defend their favored position. A Democrat, Kristy, reads editorials in The New York Times because they present substantiated arguments that help her reflect on the issues. Similarly, Anna listens to conservative radio hosts, especially Sean Hannity, because she learns from the “educated debates,” during which the host pushes opponents to explain their positions and reveals contradictions in their arguments. Katie, a Republican, occasionally listens to Rush Limbaugh because she deems that the host provides valuable arguments. Notably she later notes: “You certainly derive justification for your views. ... Rush Limbaugh brings up points that are very strong and that I couldn’t have thought of or that I wouldn&rsquo ;t have the information to make.”
Role of Interests
What role do interests play in selectivity? The interviews offer some support for the notion of de facto selectivity, or explaining attention to opinion-reinforcing information with interests that are associated with ideology (Sears & Freedman, 1967). For example, all Republican informants, who tend to major in business, read The Wall Street Journal and blogs written by various CEO’s and entrepreneurs. Those sources also offer conservative commentary. Although the informants agree with the opinions, they state that it is not the ideological affinity but rather the economic insights that motivate exposure. Sam who regularly reads The Wall Street Journal argues: “You would have a hard time making a case that The Wall Street Journal is conservative or Republican. It is about economic and business issues that most people associate with the Republicans.” The informants thus argue that it is their interests that encourage exposure, and that the tendency to agree with the source’s views is not motivated by seeking support.
More importantly, the interviews show that once individual opinions on specialized issues are formulated, the tendency to learn more about those issues from consonant sources increases. But when a novel topic arises, the informants also turn to dissonant sources, an issue addressed later. That is, selective exposure seems to increase when interests are specialized and knowledge is extensive. For example a 21-year old liberal activist, Neal, is particularly interested in organized labor, gay rights, and abortion rights. He subscribes to The Nation and reads Harper’s and The Atlantic, because “the alternative media ... will highlight the stories that appeal to someone like me. So when I go to the Nation they often have a story that says about the anti-labor practice that is going on in America.” Neal also visits a liberal blog, dailykos.com because:
A 20-year old Democrat, Daniel, who says “I am a partisan more than I am an ideological person,” states that social issues are not crucial for his party involvement. “The things that really interest me are related to policy and elections,” such as raising funds, voter outreach, and policy implementation. To learn about these issues, he reads a Democratic blog mydd.com, which is “more party focused than ideologically focused, … raises money, rallies for candidates, and ... does ad watch compilations.” A Republican, Sam, emphasizes: “I am always for a very limited government and I’m always very skeptical about government involvement in certain things that should be left out to the free market,” turns to online reports on specific domestic and economic policies compiled by the conservative Heritage Foundation or the libertarian Cato Institute.
Do informants avoid dissonant media? The few who do seem to be motivated not by the discomfort of having their views challenged. Rather, they simply dismiss certain outlets. This dismissal may stem from comfort-motivated partisan avoidance; an issue addressed in the discussion. First, certain sources are avoided because they are perceived as unworthy and their coverage as boring or unsubstantiated. Chris, a Democrat who reads The Wall Street Journal, says: “There is absolutely never a time that I watch Fox news. ... I think it’s just crap.” He would also not listen to conservative talk radio, because “They are uninformed and just wrong.” Katie despises sources she classifies as liberal. She says: “I absolutely hate just how the media and the CNN have liberal viewpoints.” She also finds it “ disgusting that they are publishing classified information and giving our secrets to the enemy,” and sees it as “irresponsible journalism and the biggest threat to national security ever.” Similarly Mike does not read The New York Times or listen to NPR because they are “extremely boring” and “whining most of the time.” He says: “I don’t think that people enjoy listening to a lot of people complaining all the time. And liberals just tend to complain.” In turn Joe, a Democrat, would never watch Fox because he sees it as “ranting and raving,” and employing an “apocalyptic scare tactic BS. ‘Our freedom is in danger our country is in danger.’” He sees such coverage as “appealing to people’s emotions rather than brains,” lacking substance, and not worthy of his time.
The informants also justify selective avoidance with their knowledge, informed opinions, and quality education. In other words, they consider themselves sufficiently familiar with dissimilar views so as to not attend to sources that present such views. Katie, a Republican, says that oppositional perspectives do not affect her and she does not need to incorporate them, because “I spent four years at Penn trying to educate myself very well.” Similarly, although Neal recognizes that people who shield themselves away from other views might be close-minded, he says: “I feel that if I have taken 18 years up to this point to formulate my opinions and if I want the media to justify some of them without lying to me, this is an acceptable thing.” Also Mike, a Republican, does not consider oppositional perspectives because “I had already heard those positions anyways, so there is nothing that I am pondering. I got to the point that I just don’t read any more.”
Importantly, those who tailor their information input nevertheless encounter opposing views. More often than not, however, this inadvertent exposure is filtered through and interpreted by attitude-consistent outlets. The liberal sources that Neal follows “pick up instances in which Fox or another media outlet would spin something in a different way.” This was the case when Rush Limbaugh claimed that Michael J. Fox faked his Parkinson’s symptoms in an ad for a Democratic Senate candidate. Neal “immediately got an email from The Nation saying ‘look what Rush did,’ ‘look what’s happening here.’” Neal says that such information often comes with interpretation and usually follows the pattern: “‘Limbaugh viciously attacked ... yesterday. He said quote x, y, z and here is why he is wrong.’”
Another noteworthy finding regarding selective avoidance pertains to the problems of classifying what constitutes “challenging perspectives” or “cross-cutting exposure.” Some informants, primarily strong Republicans, claim that they are exposed to opposing views when they turn to CNN or The New York Times or when their preferred outlets present “two sides of a story.” Mike feels that “it is really important to get both sides of the story” and hence he “would also watch normal news,” by which he means CNN or MSNBC. Several conservatives also claim to obtain the Democratic perspective through Fox’s shows, such as Hannity and Colmes, which features a program with a Democratic host, Special Report with Brit Hume, which sometimes invites “liberals from National Public Radio,” or conservative talk radio, when the hosts interview Democrats .
Motivations for Seeking Challenges
Generally, informants do not avoid dissonant outlets. This could come as a hopeful finding, as cross-cutting exposure may increase tolerance and familiarity with dissimilar positions (Mutz, 2002). Some informants indeed report that “expanding horizons” and “open-mindedness” motivate them to use attitude-inconsistent sources. Chris, a fiscally conservative Democrat, reads The Wall Street Journal because there is “nothing wrong in questioning your views.” Kristy, a Democrat, watches Fox news because such exposure “is very important, even if you don’t agree with it or have a very different base of values.” Anna, a moderate Republican, reads The New York Times to “see where things stand on the other side and how they view opinions that I might have and what they have problems with.” She says: “the negative response I get from The New York Times doesn’t make me stop reading their articles.”
Importantly, interests also encourage attention to dissimilar sources. Whereas strong views on specialized topics motivate consonant exposure, seeking news on general or novel issues appears to decrease it. Sam, a Republican, who watches Fox and visits the Heritage Foundation (http://www.heritage.org/) for specific information, turns to sources he considers liberal “for strictly political coverage.” He would thus watch C-Span, Meet the Press, or CNN election specials because “when it comes to public policy arguments I don’t like rhetoric, I want substance.” He would read The New York Times to learn about “liberal issues,” such as Nancy Pelosi endorsing John Murtha during the 2006 Midterm Election. Similarly several Democrats who study business or are interested in finance read The Wall Street Journ al. For example, Sean, a liberal, reads The Economist and The Wall Street Journal because he appreciates their foreign policy coverage despite their “fiscal conservative mode.”
When probed further however, informants reveal less sanguine motivations for cross-cutting exposure. For one, they want to validate own views. Although a Democrat, Kristy first touts such exposure, she later says that regularly watching Fox “helps you validate your own beliefs. You can say that you believe in something but not until you are challenged can you actually say that you know how to defend it.” Exposure to attitude-discrepant sources might not only justify but also reinforce prior beliefs, especially when a person counterargues. Kristy says: “I often think, ‘no this is not right, because this is what I believe and think on this issue’.” Sam, a Republican who sometimes reads op-eds in The New York Times, also admits: “I would examine the validity of the arguments, form a counterargument, see where they are coming from, and where the shortages are.” Mike, a Republican, says: “When I hear the arguments I just say ‘here are the Democrats again. They have been trying to persuade me about that for 12 years. I have plenty of arguments against it.’” Chris, a Democrat who earlier said that questioning own views is desirable and reports reading the editorials in The Wall Street Journal, would nevertheless “stand up and say ‘this is not exactly right, these are my facts.’” Sean, a social liberal who reads economically conservative newspapers and magazines because he appreciates their international coverage, admits: “I generally would try to counterargue. I would try to break an argument down to small components and say what I disagree with.” As a result, he concludes that “reading them and knowing them solidifies your own arguments and views. ... Even if I can’t shoot down their arguments very well, I can definit ely strengthen my own beliefs.”
In addition, when the informants do not have sufficient knowledge to rebut some claims, they search for information that supports their position and for sources that reveal limitations in the opposing arguments. When Sean is unable to counter such arguments, he seeks more evidence. “I wouldn’t just take their stand on it but I would try to get another side of the story and learn a little bit.” Similarly, after watching An Inconvenient Truth, Al Gore’s documentary on global warming, Sam “read a Washington Journal article about this movie by an environmental science professor from MIT who talked about a lot of fallacies in that movie that was obviously biased.”
Also, the informants attend to dissimilar outlets to hone their own debating skills. Neal, a liberal, occasionally reads conservative press. “When I go to read an article on the National Review I would typically say ‘know my enemy’ rather than go there with a mindset ‘let’s see what this right wing guy has to say.’” For Neal this is useful, as it “will help me as a political activist to know the arguments they would be using against me and the arguments that they are making. So I am going to go there and read that more for that purpose than go there and read that for my understanding of an issue.”
Another motivation behind turning to dissonant sources is to detect the “spin.” Katie, a Republican who hates CNN, later says that she would visit cnn.com and foxnews.com to see how they present the same story. She admits looking for a liberal bias on CNN and claims to find it. According to her, CNN’s “sensational, emotional, capturing” language leads viewers to unconsciously adapt a democratic perspective. Neal, a liberal, would go to foxnews.com “to see how they are spinning a certain issue. ... Given that I have already read articles that say that Fox is conservative, it allows me to write it off and say that they are going to spin it that way and I am not going to believe it.”
Persuasion by Opposing Views
It is not only important to explore whether partisans attend to dissimilar sources and why, but also to assess how this exposure affects them. Due to their motivations, informants are not likely to impartially consider or objectively evaluate dissimilar arguments (Kunda, 1990). Also, due to their strong views and extensive knowledge, the informants are not likely to be persuaded (Zaller, 1992). Indeed, almost none recall being convinced by different perspectives.
Neal, The Nation reader interested in social issues, states that for him “to change an opinion is going to require a barrage of evidence to the contrary because I am pretty good at explaining away arguments that disagree with me.” Although political views that Anna holds are not as strong as Neal’s, she echoes: “For me to change my views it would require a lot of evidence since my views are pretty established.” Chris, a socially liberal reader of The Wall Street Journal, says he is “not going to read an editorial on pro-life issues or why gay marriage or universal marriage should be banned. My mind is not going to be changed.” Sam also recalls: “I have not been persuaded by a piece that would disagree with my opinions. … Something I know a lot about and have a recurrent structured way of thinking ab out, I would probably not change my opinion.”
On issues on which they do not have strong opinions and extensive knowledge, however, informants might move towards opposing perspectives (Zaller, 1992). After saying that his established opinions would not change, Sam adds: “Things I could be persuaded on are the issues I don’t generally focus on,” and also in “a situation when I don’t know much about an issue.” Also Anna would not reconsider her views “unless it is a subject that I don’t feel very strongly about or a subject that I don’t know much about.” Kristy, in fact, recalls being influenced by Fox and moving away from the Democratic party on the issues she was unfamiliar with, namely “sending American jobs oversees” and “disallowing imports from foreign countries to protect workers in the United States.”
This study aimed to put people´s voices back into the scholarship on selectivity. Drawing from in-depth and loosely structured interviews with young partisan activists, this study addressed not only the media selection stage, asking whether informants primarily turn to consonant sources or whether they also attend to oppositional outlets, but also scrutinized individual motivations behind media selection and reactions to attitude-consonant and dissonant information. On the most general level, this study shows that the interviewed partisans select some news media that confirm their predilections without avoiding dissonant sources, and do not see their choices as motivated by cognitive dissonance. Although these notions are not new, this study confirms them using an approach not often employed in selectivity research. This confirmationadds to the existing scholarship and also restores the audience member to “his rightful place in the dynamic, rather than leavi ng him in the passive, almost inert, role to which many older studies relegated him” (Klapper, 1963, p. 527).
This study also complements and extends the scholarship by offering more nuanced findings. These findings, however, are subject to several limitations, which are primarily due to self-presentation and selectivity. As already mentioned, it is unclear whether people are able to access and reliably report their experiences related to media use. Social scientists who rely on self-reported data have long wondered what meaning should be attached to individual testimony regarding his or her attitudes, behaviors, or internal states, fearing that these testimonies could be inaccurate or misleading (see Miller et al., 1993). Some scholars have specifically criticized audience research for assuming that self-reports can validly determine motives, noting that this assumption may be “a little simplistic or naive” (Severin & Tankard, 1997, p. 335) because people may lack access to the emotional or cogn itive processes that guide their behavior or may be unwilling to report their motivations and reactions to media content (Rosenstein & Grant, 1997). These critiques may well extend to this study. Also, due to social desirability biases, people may not admit that their preference for some sources results from political affinity. Well-educated partisan activists might be yet more likely to present their media diet in a light they deem appropriate when interviewed by a scholar from the same university.
At the same time, this study presumed that people can access and are willing to report their internal states (Hermans & Van Snippenburg, 1996; Renckstorf, 1996; Stromer-Galley, 2002). Even if this access and willingness may sometimes be limited, the long interviews were particularly useful in overcoming these limitations. As the interviews progressed, the differentiation between “front stage-back stage” presentational styles seemed to diminish (Goffman, 1974). Informants appeared to become more open and comfortable, increasingly relying on a colloquial language, inserting inappropriate vocabulary, and revealing information that could be regarded as threatening to their self image, such as some socially undesirable motives for and gratifications from counter-attitudinal exposure (see also Wenner, 1983 ). Further, the grounded theory approach and the process of fragmenting and reassembling the textual data allows researchers to identify and elaborate on the informants’ motivations and gratifications despite the potential limitations in the informants’ reports (see Glaser & Strauss, 1967).
Nevertheless, research relying on other methods is needed to validate and complement the findings present here. Journals and on-site observations could supply the material for “thick descriptions,” offering insight into sense-making processes and portraying the audience interactions with the media (Bennett, 2001). Quantitative studies, in turn, could validate informants’ reports by measuring emotions or cognitions during or immediately after consonant or dissonant media exposure, for example testing affect with physiological reactions and cognitive engagement with argument recall.
In addition, qualitative research involves selectivity on all levels. This study presents media consumption practices among a dozen students from an Ivy League university. This small sample is also demographically constrained and does not represent party activists from non-academic circles, partisans who do not belong to organizations or habitual voters who vote for their preferred party candidate. Interviewing those people would reveal information about selectivity that is naturally absent from the presented findings. On the analytic level, the 22 hours of recorded discussion are filled with interesting, yet unrelated, idiosyncrasies. During the analysis, these idiosyncrasies had to be reduced, excluding information that does not directly pertain to the topic at hand (see Glaser & Strauss, 1967).
Despite these limitations, this study offers several noteworthy findings and suggests potential bridges between research on selectivity and the uses and gratifications theory. First, selective exposure among the interviewed partisans is an acknowledged and motivated phenomenon primarily with regard to non-mainstream sources that are explicit about their ideological affiliations. Determining whether it takes place with mainstream news presents problems that research on selectivity should address in order to assure that this foundational concept is uniformly understood and methodologically established. For example, who should be the one who categorizes a leaning of a particular source? The active selective exposure to alternative sources is in line with the notion underlying uses and gratifications, namely that “media selection and use, is goal-directed, purposive, and motivated” (Rubin, 1994, p. 420), while the mainstream media use might serve to simply fulfill surveillance needs, without being related to belief reinforcement.
Second, non-political motivations emerge as central to explaining selectivity. Informants’ media choices do not appear to be driven by a desire to obtain psychological comfort or protect prior views. Instead, selective exposure seems to result from socialization, seeking enjoyment, and a desire to expand knowledge. Selective avoidance, in turn, is attributed to seeing certain sources as boring, unworthy, or unsubstantiated. Although not in the partisan-selectivity context, these factors have been explored within the uses and gratifications research, which suggests that media use is related to socialization and the social environment (Rosengren, 1974) and that people avoid certain content because they do not believe the message or expect to be bored, not to protect their beliefs (Blumer & McQuail, 1969). This notable finding merits elaboration, as these ostensibly non-political motivations may be related to partisanship.
With regard to the socialization-driven selectivity, through frequent interactions, normative influence and affective bonds families shape not only members’ party attachments but also media use patterns (Lazarsfeld et al., 1944; Liebes & Ribak, 1992). Attitudes developed during socialization would thus make a person more inclined to rely on certain sources and create partisan-related expectations and evaluations regarding media content. Partly speaking to these notions, research on hostile media effect finds that generalized beliefs about media bias – likely influenced by a family – lead partisans to perceive bias in specific news reports (e.g., Schmitt, Gunther, & Liebhart, 2004), and that like-minded talk, which often occurs with close ties, reinforces perceptions among partisans that some sources are biased against thei r views (Eveland & Shah, 2003). Ultimately, these processes would lead a person to rely on consonant outlets and dismiss dissonant ones.
Such motivation for selective exposure as seeking enjoyment and such factors behind selective avoidance as boredom may also be linked to partisanship. Specifically, people tend to react with positive affect (e.g., enthusiasm), to messages that are consistent with their expectations, and inconsistent messages trigger negative affect (e.g., irritation) (e.g., Marcus, Neuman, & MacKuen, 2000; Munro & Ditto, 1997). The interviewed partisans could thus experience positive emotions when tuning to attitude consonant outlets, thus perceiving them as enjoyable and entertaining; and the opposite for dissonant outlets. Ultimately, the informants would adjust their exposure so that to obtain or sustain the positive states and avoid the negative ones, although the partisan motivations would be neither salient nor acknowledged.
Further, partisanship may also underlie the other identified motivations for selectivity, a desire to learn and a perception that certain sources are unsubstantiated. Research on information processing finds that partisans see attitude-consonant messages as “fair” and “factual,” regard inconsonant messages as “unfair” and “false” (e.g., Eagly & Chaiken, 1993), perceive desirable political arguments as more likely to be true, and evaluate attitude-confirming messages as stronger than disconfirming arguments (Munro et al., 2002; Taber & Lodge, 2006). More germane here, hostile media effect scholarship also shows that partisans tend to see mainstream news sources as biased or inaccurate (Tsfati, 2003) and to perceive a specific media message as accurate if it supports their positions and inaccurat e when it opposes them (Schmitt et al., 2004). Inasmuch as perceiving a source as informed stems from agreeing with the source, the informants would explain their reliance on certain outlets with the motivation to learn, without fully realizing the political basis of this motivation. In a similar vein, because people discredit the media messages they distrust (Wober & Gunter, 1988) or judge such messages as not true (Cozzens & Contractor, 1987), the informants would a priori reject some sources for reasons that are politically driven, although not related to immediate psychological discomfort.
Because the described processes are closely intertwined, this study cannot disentangle the “unique contribution” made to media exposure by such non-political motivations as excitement seeking from such partisan motives as discomfort avoidance. The revealed nuances may invite scholars to further look into these issues, closely exploring non-political factors predicting selectivity, and especially their possible interconnections with partisanship and individual psychological comfort or desire to have own views confirmed.
This study also speaks to the role played by interests. It suggests that partisans with formulated views and extensive information on specialized issues expand their knowledge by turning to consonant sources. When motivated to learn about novel, general, or non-salient topics, the interviewed partisans use counter-attitudinal outlets. In other words, it may not be partisanship strength but rather interest strength, issue specialization, and topic-specific knowledge that exacerbate or mitigate selectivity.1 This finding is consistent with recent studies, according to which utility impacts whether people avoid or seek out dissonant messages (see Valentino et al., 2009) and with a notion, stemming from uses and gratifications, that people use content, oppositional or not, because it is functional (Grunig, 1979). Although scholars increasingly atte nd to interests, issue relevance and information utility (Valentino et al., 2009), selectivity literature would benefit from more studies that distinguish the conditions in which partisans prefer the comforting reinforcement from those in which curiosity leads to counter-attitudinal exposure.
The final notable finding offered by this study regards the motivations behind seeking challenges. Informants generally do not avoid the outlets they see as oppositional. This is because challenging information can be useful. This is especially true for activists, who need to know what their opponents think, what arguments those opponents use, and how those arguments can be rebutted. As a result, the informants do not impartially evaluate dissimilar arguments, nor do they use those arguments to reconsider own predilections. Rather, the motivations that those partisans report, such as validating own position, detecting “spin,” and honing own debating skills, might reinforce their prior views.
This finding has important implications. When open-mindedness is regarded highly and when social desirability biases would lead people to report that counter-attitudinal content expands their worldview, young and educated partisans face a researcher and admit that they use dissonant media to buttress their views. This finding also invites questions about evaluating selectivity. Although this phenomenon has long been regarded as socially disadvantageous (see Mill, 1859/1991), selectivity research has not fully addressed how users comprehend and evaluate congenial and challenging content. Also, whereas deliberative theorists have hoped that cross-cutting exposure fosters understanding, encourages people to take opposing views into account, and results in “[s]ound political judgment” (Page, 1996, p. 2), research on motivated reasoning and biased processing suggests that such exposure may sometimes backfire. N ot only does the motivation with which people attend to information affect reasoning, in that partisan motivation “may cause people to make self-serving attributions and permit them to believe what they want to believe because they want to believe it” (Kunda, 1990, p. 480). Also, strong opinions lead to biased processing, in that people who are committed to their position counterargue and discredit dissonant evidence, while concurrently producing reasons that support their position (e.g., Munro et al., 2002). As a result, exposure to dissimilar views may increase attitude polarization (e.g., Meffert et al., 2006; Tormala & Petty, 2002; Wojcieszak & Price, 2010) and exacerbate conflicts (Mendelberg & Oleske, 2000). This study echoes these preoccupations, r evealing their validity via in-depth interviews.
On a final note, this study shows that synthesizing traditional selective exposure research with the uses and gratifications theory and methods can broaden what we know about political biases at play during media content selection and evaluation. Addressing media users’ motives behind the increasingly complex media activities and the gratifications derived from selecting or avoiding certain content sheds important light on what occurs before or after the media choice and also reveals additional factors that affect selective exposure and avoidance and which are usually overlooked by survey-based selectivity research.
All in all, this study poses some questions, uncovers under-analyzed patterns, and suggests that uses and gratifications can explain some phenomena related to selectivity and offer new venues for research. More importantly, this study suggests that determining that selectivity happens tells us relatively little. It is hearing how media users articulate their experiences that reveals the meanings and the interpretations that underlie media consumption and that ultimately demonstrates that giving people voice might not only contextualize but also enrich the scholarship.
“The Penn Democrats engage in activities that support and promote the interests of the Democratic Party while trying to increase political awareness, participation, and activism on campus. We have campaigned for candidates running for local, state, and national offices. We meet regularly, conduct voter registration drives, work to improve voter turnout, engage in political discourse and debate, and bring in speakers to further the discussion of relevant issues. In addition, because we believe that it is crucial not only to promote our beliefs but also to act upon them, community service is an important part of our organization.” Some of the organization’s events include canvassing for a Democratic congressional candidate, hosting Democracy for America's Training Academy, where “progressive activists from across the country gathered to be trained in grassroots organizing,” meeting with Democratic politicians, attending the Young Democra ts of America conference, and participating in on-campus debates.
“We are an organization of students concerned about our country and the people that lead it. We strive to promote limited government, greater economic liberties for our citizen, and a general love for our country. We have over 1000 students on our listserve and many whom are active in our organization. We have accomplished much in just the last year and we will strive to do more. The accomplishments mentioned on the website include: “Registering over 400 people to vote during the 2004 Election; Enlisting many students to help get out the vote for President Bush during the 72 hours before the election; Promoting a Republican agenda on an extremely liberal Ivy League Campus; Creating an environment on campus where conservative students feel welcome to express their political views; Helping to build academic freedom, so that students of all political points are not scared or fear retribution for expressing their views publicly.”
Questions asked during the interviews.
Albarracín, D., & Mitchell, A. (2004). The role of defensive confidence in preference for proattitudinal information: How believing that one is strong can sometimes be a defensive weakness. Personality and Social Psychology Bulletin, 30, 1565-1584.
Bennett, T. (2001). Reception study, cultural studies, and mass communication. In James L. Machor & Philip Goldstein (Eds.) Reception study: From literary theory to cultural studies (pp. 203-212). New York and London: Routledge.
>Garrett, R. K. (2009). Echo chambers online?: Politically motivated selective exposure among Internet news users. Journal of Computer Mediated Communication, 14(2), 265-285. doi:10.1111/j.1083-6101.2009.01440.x
Iyengar, S., Hahn, K., & Prior, M. (2001, September). Has technology made attention to political campaigns more selective? An experimental study of the 2000 presidential campaign. Paper presented at the Annual Meeting of the American Political Science Association, San Francisco.
Jonas, E., Schulz-Hardt, S., Frey, D., & Thelen, N. (2001). Confirmation bias in sequential information search after preliminary decisions: An expansion of dissonance theoretical research on selective exposure to information. Journal of Personality and Social Psychology, 80, 557–571.
Katz, E., Blumler, J. & Gurevitch, M. (1974). “Utilization of mass communication by the individual.” In J.G. Blumler & E. Katz (Eds.), Uses of Mass Communication (pp. 9-34). Beverley Hills, CA: SAGE Publications.
Krosnick, J. A., Booninger, D. A., Chuang, Y. C., Berent, M. K. & Carnot, C. G. (1993). Attitude strength: One construct or many related constructs? Journal of Personality and Social Psychology, 63(6), 1132-1151.
LaBennet, O. (2006). Reading Buffy and ‘looking proper’: Race, gender and consumption among West Indian girls in Brooklyn. In Kamari Clarke & Deborah A. Thomas (Eds.), Globalization and Race: Transformations in the Cultural Production of Blackness (pp. 279-298). Durham: Duke University Press.
Miller, A., McHoskey, J., Bane, C., & Dowd, T. (1993). The attitude polarization phenomenon - role of response measure, attitude extremity, and behavioral consequences of reported attitude change. Journal of Personality and Social Psychology, 64, 561-574.
Munro, G. D., & Ditto, P. H. (1997). Biased assimilation, attitude polarization, and affect in reactions to stereotype-relevant scientific information. Personality and Social Psychology Bulletin, 23, 636–653.
Munro, G. D., Ditto, P. H., Lockhart, L. K, Fagerlin, A., Gready, M., & Peterson, E. (2002). Biased assimilation of sociopolitical arguments: Evaluating the 1996 U.S. presidential debate. Basic and Applied Social Psychology, 24, 15–26.
Renckstorf, K. (1996). Media use as social action: A theoretical perspective. In K. Renckstorf, D. McQuail & N. Jankowski (Eds.), Media use as social action: A European approach to audience studies (pp. 18-31). London, ENG: Libbey & Co.
Rosengren, K. E. (1974). Uses and gratifications: A paradigm outlined. In J. G. Blumler & E. Katz (Eds.), The uses of mass communications: Current perspectives on gratifications research (pp. 269–286). Beverly Hills, CA: Sage.
Rubin, A. M. (1994). Media uses and effects: A uses and gratifications perspective. In J. Bryant & D. Zillmann (Eds.), Media effects: Advances in theory and research (pp. 417–436). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Spitulnik, D. (2002). Mobile machines and fluid audiences: Rethinking reception through Zambian radio culture. In Ginsburg, F., L. Abu-Lughod, B. Larkin (Eds.), Media worlds: Anthropology on new terrain. Berkeley, CA, University of California Press: 337-354.
Vallone , R., Ross, L., & Lepper, M. (1985). The hostile media phenomenon: Biased perception and perception of bias in coverage of the Beirut massacre. Journal of Personality and Social Psychology, 49, 577-585.
1 It is worth noting that interest strength and issue knowledge are correlated with opinion strength and its underlying components (e.g., attitude extremity, opinion intensity, attitude certainty, and so forth) (Krosnick et al. 1993), which in turn affect selectivity. In addition, issue involvement and partisanship strength are related with each other among this sample, making it difficult to firmly differentiate between interest and political-based selectivity.
Copyright 2010 Communication Institute for Online Scholarship, Inc.
This file may not be publicly distributed or reproduced without written permission of