An editorial in the Columbia Journalism Review commenting on the nation and the media after the 2004 presidential election suggests it is possible that media outlets are getting a bad rap when it comes to claims of media bias. Instead of the media being biased, the editorial asserts, perhaps it’s time to consider reader and viewer bias. “We remain a red and blue nation,” the editorial points out, adding that media outlets, driven by economics, may have to cater to audiences’ political preferences and return to the time of the partisan press when readers knew before they picked up their paper in the morning what the writers’ opinions would be (“Red News”).
Claims of media bias pervade the airwaves. Fox News often takes a right-of center stance in reporting the news but claims that it is justified in calling itself “Fair and Balanced” because of what it describes as a tendency for the rest of the mainstream news media to lean to the left (Collins & Hickey). Mainstream news outlets take a beating every week on radio talk shows hosted by the likes of Rush Limbaugh and Bill O’Reilly, who accuse the news outlets of having a left-of-center political agenda. Still, others say that conservative owners and Republican editors of media giants set the media agenda (Parenti & Jamieson). Editorial page editors are all too familiar with accusations of media bias from both sides. Schmitt and colleagues point out a common problem for newspaper editors. “Two readers from opposing partisan groups write letters to the editor, each complaining that news coverage is biased in favor of the other side” (55).
Many of those looking to back up their claims of media bias point to a 1996 Freedom Forum/Roper Center survey as evidence that the news media lean one way or the other. The survey found that out of 139 Washington-based journalists surveyed, 61 percent of the sample identified themselves as “liberal” or “liberal to moderate.” Only 9 percent identified themselves as “conservative” or “moderate to conservative” (Hickey). The book The American Journalist reported in 1996 that 44 percent of journalists lean toward the Democratic Party while 16 percent prefer Republicans and 16 percent prefer independents (Dennis). In his best-selling book Bias, journalist Bernard Goldberg lays out a number of studies that show journalists lean toward the left and contends journalists as a group are generally out-of-sync with “mainstream America” (127).
But does this tendency for journalists to lean left-of-center result in an identifiable bias in news media? A sampling of content analyses show it does not. A study done analyzing news coverage during the presidential campaigns of 1988, 1992, and 1996 found that on the whole, campaign coverage of the Democratic and Republican candidates was evenly balanced in two of the three elections — the one aberration was in the 1992 campaign: Bill Clinton was found to receive 54.4 percent of positive coverage in his race against George H.W. Bush (Watts, Domke, Shah & Fan). A study analyzing news coverage of the 1991 House banking scandal, in which members had written more than 8,000 checks resulting in insufficient funds in their House bank accounts, found that coverage was evenly balanced between Democrats and Republicans, showing no evidence to support claims of “anti-Republican bias or any meaningful partisan bias” (Niven 648).
As researchers continue to undertake content analyses in response to media bias claims, another body of research is ongoing. This research is looking at audience perceptions and asking what consumers of news and information bring to the table when it comes to seeing a news story as biased. Could the news stories some readers and viewers deem biased actually be measured as fair and objective by other readers and viewers? Could it be that audience members’ preconceived notions are responsible for the way they judge stories, and not the stories themselves? Though the answer is not that simple, previous research suggests individuals’ preconceived notions do play a big role in how they judge the fairness and objectivity of a news story.
This study expands on that body of research, looking at audience perception of media bias and the hostile media effect, which Vallone and colleagues first described in 1985 as the tendency of partisans on both sides of an issue to see new coverage as biased against their point of view. Since Vallone and colleagues’ coinage of the term, many researchers have sought to replicate the hostile media effect, and many have been successful, finding that two viewers diametrically opposed to each other on an issue covered in the news will find the same story biased against their point of view. Issues covered in these studies have included the ongoing Arab-Israeli conflict (Vallone et al.), the use of primates in research (Gunther et al.), and the use of genetically modified foods (Schmitt et al.).
This study attempts to replicate the hostile media effect by looking at how those who support and oppose the war in Iraq see a news story related to the war. It attempts to expand on the knowledge of the hostile media effect by seeing if it is strong enough to drive readers and viewers to other sources they see as friendly to their point of view.
Why is it important to better understand the hostile media effect and its implications? Schmitt and colleagues put it this way: “…if people do subject information to different cognitive and perceptual processing when it appears in mass media, understanding this processing may provide us with important clues to the singular role of mass media as an information source in society” (55).
It is important to study this effect and understand what impact it will have in a society that is growing more segmented when it comes to media consumption. Evidence shows that audiences are becoming more fragmented based on political beliefs and other characteristics. Mainstream media outlets are losing viewers and listeners to new media sources, including cable television programming, Internet Web sites, chat rooms, and talk radio. Jones states “For any given new media program, members of the audience are more likely to share common interests and political views” (159).
Looking at the hostile media effect and the role it plays in selection of news sources sets the stage for answering further important questions: How will getting news and information from a source that presents it in a light friendly to one’s political outlook affect their knowledge on issues? How will it affect political discourse and one’s willingness to engage in discussions with those who share differing views? How will citizens get critical information about elected officials and issues when media outlets serve as an extension of the political party and promote one side of the debate while discrediting the other?
Review of Literature
Trends in media consumption
In 1996, Australian Rupert Murdoch launched Fox News in an attempt to be a major player in American television (Hickey). Nine years after its launch, Fox’s opinion-based format with a host of news personalities who boast mostly right of center views has helped the network eclipse the viewership of CNN and drawn people away from network news to cable (Pew). Fox continues to climb ahead of CNN in the Nielsen ratings. On election night 2004, Fox drew 8.1 million viewers compared to CNN’s 6.2 million viewers. Fox beat even the broadcast networks in viewership during the Republican convention in September 2004 (Berkowitz).
Results from a 2004 Pew Research Center for the People and the Press survey show that since 2000, the number of Americans who watch Fox News has increased from 17 percent to 25 percent (Pew). Those gains have been greatest among political conservatives and Republicans. Fifty-two (52) percent of Fox viewersdescribe themselves as conservative, up from 40 percent when the survey was taken in 2000. While Republicans and political conservatives are coming to Fox in larger numbers, they are also shunning other news sources. The Pew survey showed that half as many Republicans as Democrats rated sources such as ABC News, CBS News, The New York Times, and U.S. News and World Report as credible.
The partisan divide isn’t present only in Fox viewers. The gap between the percentage of Republicans and Democrats who listen to talk radio is also growing, according to the Pew study. It defines talk radio as shows that invite viewers to discuss current events, public issues and politics and includes examples such as Rush Limbaugh’s show and call-in shows on National Public Radio. In 2004, 24 percent of Republicans regularly listened to radio shows that invite discussion of current events and politics. Only 13 percent of Democrats listened to these shows. In 2002, 21 percent of Republicans listened to talk radio versus 16 percent for Democrats.
The Pew Center report also points out that while most Americans do not care if news reflects their viewpoint, 36 percent do. Thirty-three (33) percent of those who identified themselves as liberal reported liking news that shares their point of view. The report says liberals go to sources such as Comedy Central’s “The Daily Show” and National Public Radio to get news and information. Forty-three (43) percent of conservatives reported liking their news to share their point of view and 41 percent of Republicans go to Fox News to get news that coincides with that view. Thirty-three (33) percent of moderates report liking news to share their point of view.
Evidence points to a growing trend: Political partisans sense a growing presence of bias in mainstream news. This perception of bias leads them to seek out news programming and publications that square with the views they hold. The success of opinion-based programming shows that when given a choice, partisans will often abandon other news sources and stick with those they feel present the news in the light they most want to see it. The rise of cable and the Internet has given readers and viewers unprecedented choice in deciding where they want to get their news and information. This level of choice threatens to fragment audiences in new ways.
It is important to look at what drives the perception of bias, particularly among partisans. As the Pew Center data show, media bias is driving viewers and readers — particularly Republicans —to other specific sources of news and information that they and like-minded news consumers deem friendlier to their views. Building a theoretical framework: The hostile media effect Vallone and colleagues were the first to coin the phrase “hostile media phenomenon” to describe “the tendency for partisans (defined in this case as those who share opposing views on an issue) to view media coverage of controversial events as unfairly biased and hostile to the position they advocate” (Discussion section 1). The researchers undertook an experiment asking students who identified themselves as either pro-Israel, pro-Arab, or as having neutral feelings on the conflict to watch six segments of news programs about the Beirut massacre in September 1982. When it came to the pro-Israeli and pro-Arab students, each side found the news segments as biased in favor of the other side.
Further, the researchers found that the perception of bias led those on both sides to believe that personal views of the editorial staffs who produced the segments were opposite of their own when it came to the subject. Vallone and colleagues also noted that within both groups, those with the highest amount of knowledge were more likely to identify the media as biased. In their study, the researchers wrote, “Our suggestion is that viewers of media coverage of a continuing political struggle may see their side portrayed as villain, and the other side as the hero” (Media Coverage section 4). They came up with three possible explanations of why those who claimed media bias did. The first was that partisans would observe a story as devoting an equal amount of facts and time to both sides, but because partisans believed their side’s claims to be more accurate than the other, even balanced coverage would be seen as biased. Giner-Sorolla and Chaiken explain this in a later study, saying “Thus, even a correctly perceived balanced presentation would appear to be unfairly biased toward the opposition, by treating the ‘inferior’ claims of the opposition as equivalent to the ‘superior’ claims of the partisan’s own side” (166).
The second explanation was that individuals recall the same images and arguments in a news report, but they are likely to classify those images and argumentsas hostile to their own side. Even neutral material is likely to be rejected, so that thevalence of images and arguments are seen as supporting the other side. Giner-Sorolla and Chaiken call this “selective categorization.” The third explanation was that partisans remember the facts and images that went against their beliefs better than they remembered the material that coincided with their beliefs. Giner-Sorolla and Chaiken call this “selective recall.”
Seeking to further understand how processes contributing to the hostile media effect work, Giner-Sorolla and Chaiken undertook additional experiments to see if news stories on developments in the Middle East and abortion prompted thesame reaction among partisans. They were able to replicate the hostile media effect among a group of students on the Middle East issue, but not on the abortion issue. The two found through their experiment that selective recall and selective categorization did not explain the hostile media effect, but that “prior beliefs about media bias were the main ingredient on bias judgment, affecting viewers of the program both directly and through estimates of program content” (Giner-Sorolla & Chaiken 179).
Gunther built on the work of Vallone and colleagues by looking more closely at who was vulnerable to the hostile media effect. After analyzing data from a study commissioned by the American Society of Newspaper Editors (ASNE), he found that if a person identified with a particular group, he or she was more likely to say media coverage of that group was unfavorable. While Republicans judged coverage of their own group as “not favorable enough,” Democrats found coverage of Republicans “too favorable.” Gunther concluded that high involvement in an issue “increases the likelihood that a person will take a skeptical view of the source of that content” (161). Those likely to have polarized attitudes about certain issues are those who have a sense of involvement with a group that is associated with an “enduring” sense of self. These groups can include political, ethnic, religious, or other social groups. Those who identified themselves with particular groups were much more likely to say the media gave unfair coverage of that particular group.
Gunther’s research gives merit to what Joseph Klapper wrote in Public Opinion Quarterly in 1957. In looking at the effect media had on a listener or viewer, Klapper called mass media a contributory agent in a process of reinforcement — or that the media were more likely to reinforce pre-existing beliefs than to change them. Klapper discussed the power of group membership when it comes to mediating information from media, and he said that communications that were opposed to the norms of a group an audience member belonged to would be rejected, especially if the audience member highly valued his or her group membership.
Thus, he says, “The typical audience member selects from the media’s varied fare those commodities which are in accord with his existing likes, and typically eschews exposure to other kinds of material” (Klapper 467). He went on to say that the audience member’s likes stem from his reference groups. Perloff alludes to this as well in one of his studies, saying that when individuals are strongly committed to a position, “they assimilate or find support in evidence and information other people regard as ambiguous, inconclusive, and problem ridden” (238).
Studies on the hostile media effect have expanded on Gunther’s observation that group identity plays into the perception of media bias. Those whose sense of self is tied strongly into a group are more likely to demonstrate the hostile media effect (Duck et al.). Matheson built on this research with an experiment that asked Bosnian Serbs and Muslim partisans and a neutral group to respond to the coverage of the 1994 Sarajevo market bombing. Matheson found that “partisans’ perceptions of who is the villain and who is the victim stem from the motivation to distinguish between the in-group and out-group in such a way as to reinforce a positive and distinct in-group identity” (Matheson 123).
After the seminal studies on the hostile media effect were published, Gunther and colleagues investigated the larger effect of the phenomenon when it came to partisans estimating larger public opinion. Christen, Kannaovakun and Gunther stated that “hostile media perceptions are important only to the extent that they result in related attitude or behavior outcomes…” (425). Looking at the 1997 United Parcel Service strike, the researchers sought to find out whether groups on both sides of the strike — UPS workers and Teamsters — would exhibit the hostile media effect and if they did, how it would affect their perceptions of public opinion.
They predicted it could go in two directions. First, partisans who found media to be hostile to their point of view could perceive public opinion to follow the mediac overage they believed others were being exposed to (the persuasive press inference). Or, those who perceived media coverage to be biased could perceive public opinion to be a reflection of their own views (projection). The hostile media effect was replicated, with both the UPS and Teamster partisans seeing news coverage as biased against their own groups. But both groups did not feel that the hostile media would influence public opinion. Instead, both groups perceived public opinion to be a reflection of their own views (Christen et al.).
Thus, it seems that this study adds fuel to the fire of those making claims of media bias. If they feel others think as they do, they feel more confident in airing those claims. Schmitt and colleagues write about the power of partisan opinions. “Their viewpoints can powerfully influence public opinion and public policy; they are the ones who campaign and lobby, who demonstrate, parade and picket, who promote their viewpoint in countless ways…” (623-624). Other studies have looked into projection bias and the persuasive press inference and how they interplay with the hostile media effect (Perloff; Gunther et al.; Gunther & Christen).
Recent studies have looked at how sources affect the perception of bias in partisans. Arpan and Raney carried out a study to find out whether news source resulted in the hostile media effect in partisans. They asked research subjects to read a story about their hometown college football team in one of three newspapers: the hometown newspaper, the newspaper covering the city of a rival team and a newspaper from a neutral town. The hostile media effect was replicated, with fans perceiving the article in all three newspapers as more biased against the home team than the rival team. The researchers also found that participants rated the story as less biased against their team when it was printed in the hometown newspaper than when it was printed in the rival-town or neutral-town newspapers. Schmitt and colleagues found that partisans who read information about genetically modified foods in a newspaper story found the information to be disagreeably biased. But the hostile media effect disappeared when the identical information was presented to partisans in the form of a college student’s essay.
Factors influencing audience perception
Watts and colleagues write that it is “primarily citizens’ partisan views that color perceptions of media fairness rather than assessments of actual bias” (147). But they acknowledge there are other factors at work when it comes to a reader, viewer, or listener coming up with an assessment that a story is biased. Along with their preconceived notions, external factors contribute to the notion that the media could be “hostile” to a partisan’s point of view.
One group to blame is the media itself, Watts and colleagues said. They said there is a new focus by journalists, candidates, and political pundits to look at the role the news media play in elections. This has led to increased coverage of news programs and journalists themselves, something the researchers refer to as “self coverage.” A study of such coverage found it focuses overwhelmingly on the alleged presence of a left-leaning bias. Watts and colleagues concluded: “These results, then, suggest that the rising public perception of a liberal news media may have more to do with news self-coverage than with biases in valence news content” (Watts et al. 159). They point out that the rise in media self-coverage is proportional to the increase in the number of media outlets in the past 15 years.
Another factor that Watts and colleagues studied was the public’s taking cues from conservative elites who allege a liberal media bias, perpetuating a notion that mainstream news channels use various techniques to advance a left-leaning agenda. They cite poll data that shows an increase in complaints by the public during the 1988, 1992, and 1996 presidential campaigns that news coverage in general was biased in favor of Democrats. The percentage of the public claiming the press treated the Republican candidate unfairly increased 13 percent during the 1988 campaign, 22 percent during the 1992 campaign, and 9 percent during the 1996 campaign. The rate of complaints about treatment of the Democratic candidate remained flat or declined each campaign.
Watts and colleagues say studies of news coverage don’t support claims that Democrats were treated better in the press than Republicans. Rather, they say the decision that the press is liberal often results from readers and viewers taking cues from others. Because people don’t have time to pay attention to and absorb all the political news coverage that exists, they look for shortcuts. They turn to figures who, in their mind, are trustworthy to take information from and form opinions. Thus, when candidates, talk show hosts, and other conservatives lodge accusations of a liberal media bias, members of the public taking cues from them are likely to adopt similar attitudes. Watts and colleagues write, “For conservative elites, in particular, phrases such as ‘the liberal media’ and ‘the media elite’ have become embedded in contemporary political culture and may possess considerable rhetorical weight. Liberal elites, in contrast, do not seem to have a comparable phrase” (151).
The results support the findings of Giner-Sorolla and Chaiken, who said that the hostile media effect could lead a partisan to declare a news broadcast to be biased even before he saw it by applying his prior beliefs about a media outlet. Giner-Sorolla and Chaiken say: “Thus, partisans may have accepted the contention of fellow sympathizers that the mainstream media are biased” (167).
In another study built on these findings, Watts and colleagues sought to find out whether conservative elites’ claims of liberal media bias came during periods when their candidates were subject to more negative press or when they were behind in the polls. The researchers found that claims of liberal bias in the news were more likely to come about when Republican candidates were receiving positive coverage and were ahead in the polls. They suggested that the use of the liberal media accusation occurs more frequently when Republicans are able to exert more control over their message. They say this proves that complaints about news coverage are strategic to a certain point. In other words, conservatives lodge accusations that aren’t based in fact in order to exert control over a media message they have no control over.
D’Alessio furthered the contention that cues are powerful in developing perceptions of media bias. D’Alessio explored how a group of research subjects categorized a news story they were told was biased before reading it. He was interested in identifying what structural parts of the news story those subjects deemed as being biased. Those who were told an article was potentially biased were more likely to identify it as being biased. Those who were not cued were more likely to say their article was free from bias.
D’Alessio asked subjects to circle the parts of the news story they claimed contained bias and then rate the article as biased or free from bias. Though participants who were given a cue that the article was biased were more likely to use that term to describe the story, they didn’t mark more statements as biased. Thus, D’Alessio concluded that perception of bias is related to charges of bias, not actual bias, supporting the research of Watts and colleagues. “Together, their findings and ours suggest another form of cognitive simplification has been brought into play, this one related to the bandwagon effect, in which people simply ‘go along’ with majority positions” (291).
D’Alessio’s study also showed that those claiming there was bias in a news stor y were likely to point to quotes that fall outside their latitude of acceptance as making a story biased. This puts journalists in a tough position, he said, because this perception of bias results from the person who is quoted, not the journalist. It shows that simply by offering supporting quotes from each side, a journalist is opening up him or herself to accusations of bias. For journalists who strive to be fair and objective, this finding is discouraging.
Another cause of the perception of media bias to undergo examination has been the similarity of those whom partisan people tend to talk to. Eveland and Shah carried out a study predicting that partisans who undertook discussions with other partisans would perceive a greater degree of press bias in the media. They found that Republican Party identification, strong partisanship, and political involvement were positively associated with perceptions of news bias. The extent to which these groups have political conversations with like-minded people significantly increases the perception that there is widespread bias. They write, “People who frequently discuss politics with others who share their views likely have distorted standards of what constitutes unbiased media content” (113). These findings support those of Gunther and Klapper, who cited the role groups play in reinforcing ideas and sparking skepticism about the sources of media that challenge group ideas and norms.
Questions of audience perception
In his book Tilted: The Search for Media Bias, David Niven acknowledges the literature that says claims of media bias are a matter of audience perception, not of the actual content of news. But he raises questions about audience perception being the sole cause of media claims. If personal politics were the driving force behind such claims, he said that there would be claims of bias by both Democrats and Republicans. Niven states: “Both Democrats and Republicans should see massive bias in the news because of the steady stream of two-sided information in the media, half of which is sure to offend them,” (Niven 45) He does not substantiate with research his claim that information in the media is two-sided.
Karlberg and Hackett also challenge the contention that the hostile media phenomenon is in large part to blame for partisan assertions that the news is biased and that journalists will insulate themselves from allegations of bias by “playing it down the middle” and covering both sides. The two undertook a survey in Canada between respondents on opposite ends of the political spectrum (conservative/establishment and progressive/advocacy groups) asking them to identify weaknesses in media coverage. They sought to find out whether criticisms from the two groups would be mutually contradictory, which they said would be evidence of the hostile media phenomenon. The researchers found that criticisms of the two groups were not contradictory. “Both groups were critical of the media for insufficiently covering positive stores and the social implications of economic policies (Karlberg & Hackettp; Discussion section 5). They went on to say, “The key point here is the intergroup similarity and the indication that dissatisfaction with media performance cannot be reduced to perceptions of hostile ‘bias’” (Karlberg & Hackett Discussion section 5). This is one of the few studies undertaken to challenge the hostile media effect, but the study’s lack of focus on a specific issue may have failed to generate the passion and feelings necessary among specific groups of partisans as previous studies have.
Niven summarized the challenges media outlets face in today’s landscape. Whether perceptions of media bias are a result of audience perceptions or true bias in the news, the media still must respond or face a dearth of readers and viewers. He said this pressure has forced media organizations back in time. The partisan press died years ago because the demand for neutral, balanced coverage outweighed the demand for news catering to specific groups holding specific beliefs. America gave bir h to a press steeped in partisanship. Writing about the press in the late 1700s and early 1800s, Sloan and Startt state: “Clearly, editors believed the overriding purpose of the press was to serve a partisan cause” (73). They say the distinctive characteristic of the press from 1783 to 1833 was political partisanship, and they point out that newspapers played an important role during a critical stage in American politics.
During this time, as the Constitution was being ratified and America’s democracy was being shaped, readers were “annoyed” when a newspaper would not take a partisan stand and had little respect for an impartial editor. This continued throughout the 1800s and early 1900s, when circulation of large dailies was tied to political parties. But in 1920, Walter Lippmann, editor of the New York World, called for the professional training of reporters in techniques related to objectivity: attributing information to official sources, balancing coverage by telling both sides of an issue, and offering context to news events instead of presenting facts in isolation (Sloan & Startt). Objectivity has been the gold standard in journalism since then, even though public officials throughout modern history have challenged the press’s objectivity.
But today, technology is giving viewers and readers more choices. Along with the growing number of cable channels, people can find their own niche of likeminded folk and news and information catered to their interest on the Internet. Niven predicted that the popularity of conservative talk radio is the beginning of Republicans fleeing the mainstream media. He stated: “Ironically, the logical extension of the flight of conservatives from mainstream media is the rebirth of the partisan press” (Niven 49-50). This observation points to the growing potential for audiences to separate into niches based on their own preferences, and in this case, based on their political views. Webster summarized the social implications of audience fragmentation that could be sparked by partisans’ fleeing mainstream news media, saying a “worrisome prospect is that technology and advertiser-driven programming will reconfigure the mass audience into many small, relative exclusive communities of interest that never encounter dissident voices or different points of view” (366).
This study seeks to expand on previous studies done on the hostile media effect by examining whether preconceived notions about a media source make a difference in perceptions of media bias and whether they play into the hostile media phenomenon. To explore this question, an experiment was undertaken. To test for the hostile media effect, a newspaper story on a development related to the war in Iraq was presented to research subjects. The same story was made to appear in three different made-up newspapers: The Conservative Times-Herald, The Liberal Times Media Herald and The Times-Herald. Research subjects received one of the three versions.
Based on the literature, the following hypotheses have been formed.
Hypothesis 1: Partisans on both sides of the issue will rate the story that appears in the three sources as biased against their point of view based on their overall assessment of bias and how they feel each side is portrayed in the story.
Hypothesis 2: Overall, partisans on both sides of the issue will rate satisfaction with the story as low and rate themselves as less likely to turn to the source in the future to get information.
Hypothesis 3: The level of perceived bias within each group of partisans will correlate negatively with their level of satisfaction with the story and the likelihood they will return to that source for information on the issue.
Hypothesis 4: Partisans on both sides of the issue are likely to go to media they deem friendly toward their views to get more information on the war in Iraq.
Stimuli and research sample
These hypotheses will be tested through an experimental design in which participants randomly receive a news story related to the issue of the war in Iraq in one of the three stimuli mentioned above. The news story was about Cindy Sheehan, the mother of a fallen soldier who made headlines carrying out war protests at President Bush’s ranch in Crawford, Texas, and at the White House in 2005. The story is made up with quotes and information from articles that appeared in The Los Angeles Times, The San Francisco Chronicle, The Washington Post, and on the Associated Press wire the day after she was arrested for protesting in front of the White House. The stories were found by carrying out a keyword search on “Cindy Sheehan” in Lexis Nexis. The composite story is made to be as fair and balanced as possible by including information about Sheehan and anti-war protests as well as information about rallies staged in support of the administration that same weekend.
The articles are formatted to appear as if they were printed out from the newspapers’ Web sites. The names of the three stimuli are selected to answer the question: Does media source play a role in the perception of media bias? Will partisans (conservatives and liberals) deem a story biased simply based on the names of the newspapers? Made-up newspapers are chosen over existing publications because research subjects may not be familiar with the leanings the editorial pages some nationally circulated news publications are said to have.
Included in subjects’ packet will be a pre-test seeking standard demographic information as well as questions asking about party affiliation, religiosity, and how much time they spend with media each day. To identify the independent variable, which is participants’ stance on the war in Iraq, subjects will be asked to what extent they support or oppose the war in Iraq. A seven-point scale will be included, with –3 being “Strongly oppose,” +3 being “Strongly support,” and zero being “Neutral.” The scale is adapted from a similar 11-point scale that Schmitt and colleagues used in a study testing for the hostile media effect when it came to information on the issue of the development of genetically modified foods. Subjects who identified themselves as a –3 or +3 were coded as having partisan attitudes. Those who identified themselves on the scale as –2, –1, zero, –1, +1, or +2 were coded as having neutral attitudes.
Participants will be asked to read the news story in their packet and proceed to the post-test, adapted from the questionnaire Schmitt and colleagues used in their study. To measure the dependent variable, which was perceived bias, subjects will be asked to rate different aspects of the story on the same seven-point scale used above.
Questions included, “Would you say that the portrayal of the war in Iraq in this article was strictly neutral, or was it biased in favor of one side or the other?” “Would you say the portrayal of the opponents of the war in Iraq in this article was strictly neutral, or was it biased for or against them?” and “Would you say the portrayal of supporters of the war in Iraq in this article was simply neutral, or was it biased for or against them?” Two items will be included asking what percentage of the story is favorable to the issue of the war and what percentage is unfavorable. Questions will be also asked about the source the story appeared in. Subjects will be asked to rank on the seven-point scale how they felt future stories on the issue would be covered in the source they read, whether it would the The Conservative Times-Herald, The Liberal Times-Herald, or The Times-Herald.
Another section of questions was included to understand whether partisans seek out media they see as friendly toward their views in reaction to an article they perceive as biased. Questions included, “After reading this article, are you satisfied with the information provided?” and “In the future, if you wanted find out more about the war in Iraq, would you be likely to turn to this source?” The final question asked the following: “Where would you go to get the most reliable information on the latest developments in the war in Iraq? Please rank the following choices.” Sources included cable networks such as Fox News and CNN; non-mainstream news sources such as “The Daily Show,” Rush Limbaugh’s radio show, Al Franken’s radio show, the Drudge Report and the Daily Kos blog; news magazines such as The Weekly Standard and The Nation; and mainstream newspapers such as The New York Times and The Wall Street Journal.
Conducting the test
The researcher will undertake the experiment with a convenience sample: undergraduate students. A total of 30 students will be in attendance on the day the study will be administered in early January 2007. The researcher will offer no incentive to complete the survey other than to explain that students’ participation would help to better understand attitudes about news coverage of the war in Iraq. The questionnaires will be distributed randomly by teaching assistants blind to the hypotheses in class. Subjects will be given roughly 25 minutes to take complete the questionnaire. Standard debriefing procedures will be undertaken after the questionnaires will be handed in to let subjects know that they will be deceived and that none of the news sources contained in their packets exist in real life.
Expectations and Conclusions
Evidence of the hostile media effect
Both strong opponents and strong supporters would demonstrate evidence of the hostile media effect. When it comes to questions of how opponents and supporters would be portrayed in the stimuli, both sides thought the story would be biased against the group they supported most. For example, strong supporters of the war would show a much stronger perception of bias than strong opponents of the war. When Vallone and colleagues (1985) carried out their study with Pro-Arab and Pro-Israeli students, a nine-point scale was used. Partisans were clustered more toward the extreme ends of the scale than in the present study, in which partisans would be remained more likely in the center. Such results are expected in this study.
Previous studies on the hostile media effect have used partisan subjects from groups in which formal membership is required. In the Vallone study, researchers recruited members of pro-Arab and pro- Israeli student associations. Schmitt and colleagues selected partisans on the issue of development of genetically modified foods by going to annual meetings of groups with strong positions on the issue. Those who have done studies on the hostile media effect have recruited highly involved participants because the literature states that group membership is a high predictor of the hostile media effect. Christen and colleagues argue that, “While attitude extremity may be necessary, membership in a formal group (with its attendant effects on self-identity and depth of commitment) appears to be a sufficient condition for the hostile media perception” (433). As a result, they say, partisans in more general audiences would not show as strong of evidence of the hostile media effect. While targeting highly partisan groups in this way helps researcher understand more about the hostile media effect and how it plays out, the results can’t necessarily be generalized to the rest of the population.
The results of this study and other studies on the hostile media effect should give journalists some hope. While a small and vocal group are making claims of media bias, the majority of the public who take in news and information sources seem to sense that media outlets are fairly neutral when presenting the news. It is unlikely that any publisher fears losing a substantial number of readers or viewers because of claims of bias — the level of discourse on the media today indicates that people are paying attention. The rise of news and information sources, while fragmenting audiences, has the potential to draw more people into the national conversation than ever before. This expansion is the upside of segmentation — more diversity in voices.
Such diversity has the potential to lead to a better democracy. But there are dangers with audience segmentation as well. A study by Mendelsohn and Nadeau found that exposing audiences to the same group of national opinion leaders and media messages led those who initially took varying stances on issues of national importance to become less varied in their opinions. The arrival of narrowcast media, they predict, will raise number of voices, which could be a good thing, but also enhance polarization and cleavages in public opinion, which could be a bad thing. Such cleavages could result in an increase in class politics, religious divisions, and party identification.
Thus, it is important to be aware of the consequences of the hostile media effect as it relates to media choice and audience fragmentation. Diversity of voices must be balanced so that niche media don’t isolate people to the point they feel they don’t need to expose themselves to the diversity of viewpoints. While journalists can’t do much to prevent perceptions of media bias, they can acknowledge it exists and do what they can to make the public aware of it.
Arpan, L.M., & Raney, A. An Experimental Investigation of News Source and the Hostile Media Effect. Journalism and Mass Communication Quarterly, 80, (2003): 265-282.
Berkowitz, B. U.S. Election Viewers Flock to Fox News for Results. ABC News. Nov. 3, 2004. <http://abcnews.go.com/Entertainment/wireStory?id=224924>
Collins, S. Crazy Like a Fox: The Inside Story of How Fox News Beat CNN. New York: Penguin, 2004
Christen, C.T., Kannaovakun, P. & Gunther, A.C. Hostile Media Perceptions: Partisan Assessments of Press and Public during the 1997 United Parcel Service Strike. Political Communication, 19, (2002): 423-436.
D’Alessio, D. An Experimental Examination of Readers’ Perceptions of Media Bias. Journalism and Mass Communication Quarterly, 80, (2003): 282-293.
Dennis, E. How “Liberal” Are the Media, Anyway? The Continuing Conflict of Professionalism and Partisanship. The Harvard International Journal of Press/Politics, 2, (1997): 115-119.
Duck, J.M., Terry, D.J., & Hogg, M.A. Perceptions of a Media Campaign: The Role of Social Identity and the Changing Intergroup Context. Personality and Social Psychology Bulletin, 24, (1998): 3-16.
Giner-Sorolla, R., & Chaiken, S. The Causes of Hostile Media Judgments. Journal of Experimental Social Psychology, 30, (1994): 165-180.
Gunther, A.C., Christen, C.T., Liebhart, J. & Chia, C.Y.S. Congenial Public, Contrary Press, and Biased Estimates of the Climate of Opinion. Public Opinion Quarterly, 65, (2001): 295-320.
Hickey, N. Is Fox News Fair? Columbia Journalism Review, (1998, March-April): 30-36.
Jones, D. The Polarizing Effect of New Media Messages. International Journal of Public Opinion Research, 14, (2002): 158-173.
Jamieson, K.H. Everything You Think You Know About Politics…And Why You’re Wrong. New York: Basic, 2000
Karlberg, M., & Hackett, R. Canceling Each Other Out?: Interest Group Perceptions of the News Media. Canadian Journal of Communications, 21, 1996. <http://www.cjconline.ca/viewarticle.php?=id=383>
Klapper, J.T. What We Know About the Effects of Mass Communication: The Brink of Hope. The Public Opinion Quarterly, 21, (1957-58): 453-474.
Matheson, K., & Dursun, S. Social Identity Precursors to the Hostile Media Phenomenon: Partisan Perceptions of Coverage of the Bosnian Conflict. Group, Processes and Intergroup Relations 4, (2001): 116-125.
Medelsohn, M., & Nadeau, R. The Magnification and Minimization of Social Cleavages by Broadcast and Narrowcast News Media. International Journal of Public Opinion Research, 8, (1996): 374-390.
Niven, D. A Fair Test of Media Bias: Party, Race and Gender in Coverage of the 1992 House Banking Scandal. Polity, 36 (2004): 637-649.
Pew Research Center for the People and the Press. Online News Audience Larger, More Diverse, June 8, 2004 <http://people-press.org/reports/display.php3?ReportID=215>
Red News, Blue News. Columbia Journalism Review. 6, November/December, 2004
Parenti, M. The Myth of the Liberal Media. Humanist, 55, (1995, January/February): 7-9.
Perloff, R.M. Ego-Involvement and the Third Person Effect of Televised News Coverage. Communication Research, 16, (1989): 236-262.
Sloan, W.D., & Startt, J.D. The Media in America: A History. Northport AL: Vision Press, (1999).
Schmitt, K.M., Gunther, A.C., & Liebhart, J.L. Why Partisans See Mass Media as Biased. Communication Research, 31, (2004): 623-641.
Hickey, N. Is Fox News Fair? Columbia Journalism Review, (1998, March-April): 30-36.
Vallone, R.P., Ross, L., & Lepper, M.R. The Hostile Media Phenomenon: Biased Perception and Perceptions of Media in Coverage of the Beirut Massacre [Electronic version]. Journal of Personality and Social Psychology, 49, (1985): 577-585.
Watts, M.D., Domke, D., Shah, D.V., & Fan, F.P. The Politics of Conservative Elites and the “Liberal Media Argument.” Journal of Communication, 49, (1999): 35-58.
Webster, J.G. Beneath the Veneer of Fragmentation: Television Audience Polarization in a Multichannel World. Journal of Communication, 55, (2005): 366-382.