An Experimental Investigation of News Source and the Hostile Media Effect
Arpan, Laura M., Raney, Arthur A., Journalism and Mass Communication Quarterly
This study examined the interaction among different news sources, individual levels of partisanship, and the hostile media effect in sports news. Two hundred and three participants read a balanced story about their home-town college football team in one of three newspapers: the home-town, the cross-state rival university's town, or a neutral-town paper. The study found differences in the hostile media effect across conditions, suggesting the importance of news source in the phenomenon. Further, findings indicate strong support for the hostile media effect among sports news consumers.
Audience perception of credibility is considered vital to success of a given news outlet. News consumers who consider news outlets more rather than less credible tend to rely on those media, and such reliance can lead to greater media exposure.1 However, American audience evaluations of reporter and news-outlet credibility have generally been less than favorable over the past few decades. A recent poll found that, although general perceptions of media credibility had improved during the past decade, almost half of those surveyed thought the American news media usually report inaccurately, while just over half said they thought members of the press were "moral."2
The American Society of Newspaper Editors responded to rapidly eroding public trust in the news media in 1982 with a study that began to address and define credibility as it pertained to perceptions of media outlets, particularly newspapers.3 More recent research shows, however, that a news outlet's ability to be perceived as credible may stem less from reporting efforts and more from issue involvement and cognitive processes of audience members.4 Specifically, news stories intended by reporters and considered by uninvolved parties to be neutral can be perceived as biased, and therefore lacking in credibility, by highly involved news consumers. The current study used the experimental method to integrate both perspectives-source characteristics and audience involvement-to further explain the hostile media effect.
The process by which some news consumers rate ostensibly neutral stories as biased against their point of view (and/or in favor of someone else's point of view) has been termed the hostile media effect.5 Studies of the hostile media effect (HME) are informed by two classic studies of biased assimilation of information, which have shown that people with opposing views can read identical pieces of inconclusive evidence regarding an issue, yet conclude that the evidence favors their own existing opinions. One study6 found that two groups, differing in support for capital punishment and reading the same scientific statements on the efficacy of the death penalty in deterring future crime, each felt the data confirmed their prior attitudes and beliefs. Additionally, participants tended to discount the source of evidence when it was contrary to their initial beliefs and more strongly approved of the source when the evidence supported their initial beliefs. Similarly, another early study found that rival football fans who viewed the same close, penalty-ridden game came to radically different conclusions about how many infractions were assessed for each team.7
The tendency to interpret information in a manner that supports one's existing attitudes and beliefs is well documented.8 Vallone, Ross, and Lepper explain how selective perception contributes to news processing and subsequent judgments of the news media:
Partisans who have consistently processed facts and arguments in light of their preconceptions and prejudices (accepting information at face value, or subjecting it to harsh scrutiny, as a function of its congruence with these preconceptions and prejudices) are bound to believe that the preponderance of reliable, pertinent evidence favors their viewpoint. Accordingly, to the extent that the small sample of evidence and argument featured in a media presentation seems unrepresentative of this larger "population" of information, perceivers will charge bias in the presentation and will be likely to infer hostility and bias on the part of those responsible for it.9
Additionally, those who have strong opinions or who are highly involved in an issue are highly unlikely to change their minds when presented with contrary information, regardless of how compelling that information might be.10 Studies of the HME have examined how both phenomena manifest themselves in judgment of news coverage. However, contrary to early findings of biased information assimilation, most HME studies have found that people with strong opinions about the topic of a news story believe the coverage favors the other side of the issue, rather than their own opinions. For example, in a study of the Middle East conflict, pro-Israeli and pro-Arab participants viewed the same news reports about the conflict.11 Each group perceived the reports as biased in favor of the opposing side, and those most knowledgeable about the conflict were most likely to view reports as biased against their own groups.
Results similar to those in the Vallone, Ross, and Lepper study12 have been found in several other studies of the HME. Two other studies of news coverage of the Israeli-Palestinian conflict found that, again, both groups perceived the same televised news coverage as biased against their own group and in favor of the opposing group.13 Similarly, people who identified strongly with a particular social group were more likely to say that the newspaper they were most familiar with gave unfavorable coverage to that group than were people who did not strongly identify with that group.14 Another study found that when proponents of one side of an issue were exposed to news unfavorable to their side of the issue, they judged the news as more biased than did proponents of the other side.15 Participants highly involved in either side of the issue also perceived bias on the part of the reporter when the article seemed to favor the opposing side. Involvement, especially in the form of group identification, that elevates people to a "state of partisanship" is considered by some to be a moderator of the HME.16
Recent investigations of the HME have also sought to identify the mechanisms through which the effect occurs. One suggested mechanism involves judgment heuristics.17 Heuristic paths to hostile media judgments may occur in two fashions: (1) partisans may judge an ostensibly balanced presentation as hostile because such an account gives equal weight to the other side's "inferior" position or claims;18 or (2) partisans who believe the media favor the other side consistently assume bias in story content. The second mechanism represents differential processing and recall. Such differences may occur when partisans recall negative facts about their group or their side of the issue especially well, or when, though they recall the factual content of a news story equally well, partisans estimate retrospectively that a greater percentage of the information presented was unfavorable to them.19
While research on the HME has brought valuable insight into the process of evaluation of media content, only one previous examination of the HME has concurrently addressed earlier research on credibility issues associated with news sources themselves.20 Credibility beliefs can be associated not only with the news media as an industry,21 but also with individual newspapers or news shows.22 Gunther's study23 of survey participants in several social groups found no relationship between bias judgments and general characteristics of the paper (e.g., circulation size, type of ownership, political distance from own views) with which they were most familiar. However, the measure of news source effects in that study was not specific to individual papers in the final data analyses, and, as noted by Gunther, was based on characteristics associated with papers that may have been largely imperceptible to readers. Nuances or characteristics of a specific news outlet could interact with partisanship to produce a hostile media effect related to coverage of certain issues. Studies of the HME have not tested such a possibility, as all have held the news source constant or have transformed measures of individual news sources into broader variables such as general newspaper characteristics thought to contribute to perceptions of credibility.24
Purpose and Hypotheses
The goal of the current study was to integrate HME research with traditional source effects research by experimentally varying the news source and concurrently examining effects associated with individual levels of partisanship. An additional goal was to examine the HME25 in a content area (sports news) different from that studied previously (general news and public affairs). The study also examined the extent to which content-based and/or heuristic judgments seemed to be associated with hostile media perceptions of different news sources. Fans of a university football team served as the partisan participants. The study varied the location of the newspaper that printed the story by using the name and nameplate of the home-town paper, the paper from the home town of the rival team, and a "neutral" paper from another state with no teams considered rivals.
Previous studies of the HME have typically found that partisans evaluate news coverage of controversy related to their group or issue as biased against their side and in favor of the other group or side of the issue. Therefore, the following hypothesis tested for a main effect of hostile media perceptions:
H1: Participants should perceive the article in all three papers as more biased against the home team than against the rival team.
Scholars have suggested26 that elevated levels of partisanship should be associated with greater manifestations of the HME. For the current study, the level of sports fanship (particularly, team affiliation) was chosen as a measure of partisanship. While this is a prima facie departure from previous studies of the HME, the literature supports such a methodological move. Decades of sports fanship literature chronicle the psychological and sociological importance of sports team and player affiliations.27 Varying levels of sports fanship, in combination with team winning or losing, have been shown to impact public displays of team loyalty28 and disloyalty;29 self-esteem and confidence in one's physical, mental, and social skills;30 biased perceptions of a team's performance and play;31 and even perception of a U.S. victory in the Persian Gulf War.32
Though this connection has not been specifically addressed in the sports literature, findings do suggest that fanship likely functions as a form of value-relevant involvement for individuals and should, therefore, lead to the types of processing typically associated with high levels of value-relevant involvement.33 Accordingly, the following hypothesis tested for a main effect of levels of fanship on perceptions of bias:
H2: Participants with higher levels of fanship should perceive the article in all three papers as more biased against their team than should participants with lower levels of fanship.
Further, this study attempted to examine the extent to which credibility ratings of news sources might affect typical hostile media perceptions. Attribution theory34 predicts message-processing outcomes associated with perceptions of communicator credibility (or, conversely, bias). According to Eagly, Chaiken, and Wood,35 audience members receiving a message will search for cues that help them infer potential causes for the communicator's position, such as communicator attributes (e.g., prior attitudes or personality traits) and/or information about the communication situation, such as external pressures or constraints placed upon the communicator. When recipients consider communicator attributes, they likely generate expectations about whether the communicator is influenced by a knowledge bias or reporting bias. Knowledge bias assumptions are associated with beliefs that the communicator has access to limited, biased, or nonrepresentative information about the issue at hand, thus limiting his or her true ability to convey the truth. Reporting bias assumptions are associated with beliefs that a communicator's willingness to convey relevant information accurately is compromised (typically due to perceptions that the audience in the given situation will be hostile to the communicator's true position). When communicators deliver information that seems to be in line with either or both of these expected biases, audience members will likely perceive them as biased.
In the situation examined in this study, participants were likely to generate expectations about the newspapers' susceptibility to both types of bias. The home-town paper could be expected to be the only paper to have full knowledge of all the facts and (possible) mitigating circumstances related to allegations levied at the home team. Similarly, the rival-town paper could be expected to have full knowledge about the allegations levied at the rival team. The neutral paper could be perceived as having less knowledge overall and less representative information about both teams and the allegations, and thus be more subject to perceptions of knowledge bias. In this case, the participant/partisan might assume that the neutral paper reported the news in a way that reflected unfavorably on the team because the reporter likely did not have access to the full, "correct" version of the events. Such an assumption seems to be predicted by the selective-perception explanation provided by Vallone, Ross, and Lepper.36
Additionally, participants could perceive more susceptibility to reporting bias pressures at home-town and rival-town papers. They might expect audience pressure on the home-town paper to report favorably about their home team but expect the rival-town paper to feel significantly less pressure to do so. Conversely, the rival-town paper might be assumed to have more pressure to report favorably about its home team as well as pressure to report unfavorably about the home team of the participants. The latter assumption of pressure at the rival-town paper to report unfavorably about the home team of the participants certainly seems consistent with previous HME findings of correlation between prior expectations of hostility toward a particular group (e.g., Palestinians) and ratings of ostensibly neutral news stories as hostile. The neutral-town paper should be associated with weaker perceptions of reporting bias, as its audience would likely be perceived as having few supporters of either team.
The home-town paper represents the research condition in which an additive effect of knowledge bias and reporting bias should occur. Participants should expect the home-town paper to be subject to less knowledge bias and more subject to reporting bias than the other two papers. Both assumptions should result in perceptions of more favorable coverage of the home team in the home-town paper. Accordingly, the following hypothesis was tested.
H3: Participants will rate the story as less biased against their team when it is printed in the home-town newspaper than when it is printed in the rival-town and neutral-town papers.
Finally, because this study varies the source of the news while concurrently examining the HME, three research questions were posed to examine (1) whether news source is associated with differences in processing and recall of story content; and/or (2) if the source of the news changed the relationship between levels of partisanship (fanship) and perceptions of bias; and/or (3) if the source of the news changed the relationship between previous, general expectations of media bias and perceptions of bias associated with the given story.
RQ1: Do content-based evaluations vary significantly across different news sources?
RQ2: Does the relationship between fanship and perceptions of bias vary across different news sources?
RQ3: Does the relationship between general perceptions of media bias and perceptions of story bias vary across different news sources?
A group of 203 students at a large, state university in the southeastern United States participated in the research (72.4% female; 75.9% Caucasian; mean age 20.73 years). A sub-group (n=91) also completed a pretest several days prior to the main testing sessions. The sub-group was representative of the larger sample: 75.8% female; 74.7% Caucasian; and mean age of 21.15 years.
Procedures. A pretest measure of attitudes regarding general media bias was administered to a portion of the sample several days prior to the testing sessions. The main testing sessions were conducted with groups of no more than 30 participants. Next, each participant read a fictitious newspaper article; after everyone had finished reading, the researchers collected the articles. Each participant then completed a questionnaire measuring reactions to the article and fanship toward the university's sports program. Previous tests of the HME have involved at least one reference group: either a nonpartisan control group or a partisan group representing the opposite side of the given issue. However, because this study tested source effects and the HME, the researchers chose not to use a reference group, using instead a more streamlined design that might aid interpretation of results. The results presented here, therefore, attempted to illuminate possible variations in bias perceptions within one group of partisans, based upon manipulations of news source.
Stimulus Material. The stimulus newspaper article detailed recent findings of a fictitious NCAA committee that had investigated off-the-field problems with student athletes and athletic programs (e.g., crimes committed by players, recruiting violations, improper communication with sports agents). The article cited incidents occurring at two perennially top-ranked football teams in the nation: the teams from the participants' school and their major rival. The twelve-paragraph article included "excerpts" from the report listing six violations for each team, all of which were true incidents.
Three versions of the article were created appearing to have been printed in three different newspapers: the home-town paper of the participants' university, the home-town paper of the rival university, and a paper from a metropolitan area in another state. The articles were reproduced in a manner that indicated they had been clipped from each paper's morning edition.
Participants were randomly assigned to one of the three experimental conditions. Several steps were taken to reinforce the manipulation of the article's source: each version of the article included the newspaper's actual masthead; the third paragraph in the article noted that the excerpts were "of interest to us here in [home town]," "of interest to us here in [rival town]," or "concerning two of the more high-profile schools cited [home team university] and [rival team university]"; the title of the questionnaire included the title of the newspaper; and all items in the questionnaire concerning the newspaper included the paper's name.
To test the effectiveness of the manipulation, a separate sample of 44 participants with characteristics similar to the primary sample (65.9% female, 74.4% Caucasian, mean age = 19.98 years) read one version of the stimulus article and completed a shortened version of the appropriate questionnaire. As with the primary study, participants were randomly assigned to one of the three testing conditions. After completing the questionnaire, the participants were asked to identify from a list of five possibilities the city's newspaper from which the article had been clipped. Nearly all (95.5%) correctly identified the city of origin for the article.
Bias. Previous research into the HME guided the formation of the questionnaire.37 Five items ([alpha] = .72) examined perceptions of bias against the home-town team (HTT): the article was greatly biased against-in favor of HTT, the article made HTT seem unlikeable-likeable, the article made HTT seem bad-good, the writer of the article is greatly biased against-in favor of HTT, and the newspaper that printed the article is greatly biased against-in favor of HTT. These five items also examined perceptions of bias against the rival-town team ([alpha] = .75). All ten of the items utilized an 11-point scale from -5 (e.g., "greatly biased against") to +5 (e.g., "greatly biased in favor of"). Responses were averaged to yield two factors for each respondent: bias against home team and bias against rival team.
Two additional bias items forced respondents to compare how much the writer and the newspaper favored the home team versus the rival team. Both items utilized an 11-point scale from +5 (e.g., "Pro[rival team]") to +5 (e.g., "Pro-[home team]"). However, for the sake of analysis, responses on the "pro-[home team]" side of the scale were recoded to negative values (e.g., +5 became -5); as a result, the items reflected bias in favor of the rival team. The two items were significantly correlated (r = .56, p < .001). Therefore, the responses were averaged for each participant yielding a single factor: comparative bias in favor of rival team.
Recall. Participants were presented with a list of eleven statements concerning information from the article. Six statements contained information that was reported in the article; five contained unreported information. Respondents were required to correctly identify statements containing information reported in the article.38 Each item was scored as either correct (score of 1) or incorrect (0). Scores on the eleven items were summed to yield a facts correctly recalled score for each participant.
Participants were also asked to recall the number of incidents reported for each team; again, six incidents were actually listed for each team. To observe possible biased processing of information against the home team, the difference between the number of incidents recalled being reported for the home team and for the rival team was calculated. The resulting index was labeled more home-team incidents recalled.
Retrospective Estimates of Unfavorable Content. Respondents rated on an 11-point scale from 0% to 100% the percentage of the statements in the article they perceived to be favorable and unfavorable to each team. Responses were labeled percentage of statements favorable to the home team, percentage of statements favorable to the rival team, percentage of statements unfavorable to the home team, and percentage of statements unfavorable to the rival team.
General Media Bias. A six-item pretest adapted from previous research on media bias39 was administered to 91 of the participants several days prior to the main testing sessions ([alpha] = 6). Each item utilized an 11-point scale from 0 ("Don't agree at all") to 10 ("Strongly agree"). The ratings for the six items were averaged for each participant, yielding a single factor called general media bias.
Home-Team Fanship. Three items ([alpha] = .96) measured sports fanship toward the home-town team and served as an indicator of partisanship: I am a [home-team nickname] fan, I take pride in being a [home-team nickname] fan, and I want other people to know I am a [home-team nickname] fan. Each item used an 11-point scale from 0 ("Strongly disagree") to 10 ("Strongly agree"). The ratings for the three items were averaged for each participant, yielding a single factor called home-team fanship.
Random Assignment. To ensure that random assignment of participants to conditions was effective, one-way analysis of variance (ANOVA) procedures were conducted on the two independent measures. No significant differences were observed on the general media bias and home-team fanship scales.
Overall Hostile Media Effect. Consistent with the previous literature, the researchers tested for perceptions of general bias against the home team, regardless of condition. As predicted in H1, a significant difference was observed between the mean perceptions of bias against home team and perceptions of bias against rival team: t (202) = 2.86, p < .01. The paired-samples f test indicated that the participants perceived the article to be more biased against the home team (M = -1.19, s.d. = 1.27) than against the rival team (M = -.89, s.d. = 1.27). Therefore, H1 was supported.
Fanship and the Hostile Media Effect. Some scholars have suggested that higher levels of partisanship should be associated with greater manifestations of the HME. To better understand this relationship for the current project, ratings on the fanship variable were subjected to bivariate correlation analyses with the ratings on the three bias measures. Because the literature has yet to address the potential influence of different news sources on the HME, the researchers analyzed the relationship between fanship and bias without regard to the experimental conditions (n = 203). No significant correlations were observed for the three variables: fanship-bias against home team, r = -.062; fanship-bias against rival team, r = -.032; and fanship-comparative bias, r = .081. In other words, participants' levels of fanship appeared not to be significantly related to their perceptions of bias. Therefore, H2 was not supported.
Hostile Media Effect across Sources. Ratings on the bias against home team scale were subjected to a one-way ANOVA procedure by condition. A Student-Newman-Keuls (S-N-K) post-hoc comparison (see Table 1) indicated that participants in the neutral-town condition rated the article they read as significantly more biased against the home team (F^sub 2,200^ = 5.12, p < .01) than individuals in the home-town condition. Bias against the home-team ratings for participants who read the article in the rival-town paper were statistically similar to those observed in the other two conditions.
Ratings on the bias against rival team scale were likewise subjected to a one-way ANOVA procedure (with S-N-K post hoc analysis) by condition. As Table 1 indicates, a similar trend was observed: Participants who read the article in the neutral-town paper rated the article as significantly more biased against the rival team (F^sub 2,200^ = 3.48, p < .05) than individuals reading the same article in the rival-town paper. Bias against the rival team ratings for participants who read the article in the home-town paper were statistically similar to those observed in the other two conditions.
The comparative bias in favor of rival team scale forced the participants to choose whether the article was more favorable to the rival team or the home team. A one-way ANOVA procedure (with S-N-K post hoc analysis) indicated that participants reading the article in both the rival-town and the neutral-town newspaper conditions rated the article significantly more favorable to the rival team (F^sub 2,200^ = 7.45, p < .01) than participants in the home-town newspaper conditions (see Table 1). Therefore, H3 was primarily supported.
Sources, Information Recall, and Retrospective Estimates. Previous research indicated that the amount and type of information remembered in a story might play a role in the HME phenomenon. Because the HME was observed in the present study, the researchers sought to identify differences in information recall that might have led to or have been influenced by the presence of the effect. Several recall measures were included: number of facts correctly recalled, number of incidents recalled concerning the home team versus the rival team, and the percentage of statements recalled that were favorable and unfavorable to each team. Each measure was subjected to a one-way ANOVA procedure by condition. As Table 2 reports, no significant differences were observed for the variables. In other words, though the source of the story significantly impacted perceptions of bias in the story, the source did not significantly influence differences in information recalled from the story.
One exception was observed: Readers in the home-town condition tended to recall a lower percentage of statements unfavorable to the home team than their counterparts in the rival-team and neutral-town conditions. As Table 2 indicates, this finding was not significant by accepted standards (F^sub 2,200^ = 2.89, p < .10).
Fanship and the Hostile Media Effect across Sources. Previous research also indicated that an individual's level of partisanship-or in this case, home-team fanship-might play a role in the HME phenomenon. Therefore fanship ratings were subjected to bivariate correlation analyses with the ratings on the three bias measures for each condition. As Table 3 reports, no significant correlations were observed for the three variables across source conditions. In other words, a participant's level of fanship appeared not to be significantly related to his or her perceptions of bias in the story, regardless of the story's source.
General Media Bias and the Hostile Media Effect across Sources. Previous research also indicated that prior attitudes about media bias might play a role in the HME. Thus, ratings on the general media bias variable were subjected to bivariate correlation analyses with the ratings on the three bias measures for each condition. As Table 4 reports, no significant correlations were observed for the three variables across the source conditions. In other words, a participant's level of prior media bias appeared not to be significantly related to perceptions of bias in the story, regardless of the story's source.
The goals of this study were to examine concurrently source effects associated with the news outlet and the HME as well as to examine hostile media effects in the content area of sports news. As predicted by H1, participants perceived the article as more biased against the home team than against the rival team, regardless of the article's source. Based on the sports fanship literature, H2 predicted that participants with higher levels of fanship toward the home team would perceive the article as more biased against their team than participants with lower levels of fanship, regardless of the article's source. However, the data did not support H2. H3 proposed that participants would rate the story as less biased against their team when it was in the home-town newspaper than when it was in either the rival-town or neutral-town papers. The data indicated that perceptions of bias indeed varied according to the news source indicated by the experimental news story, primarily in support of H3. Perceptions of bias against the home team were always lowest when the news story was in the home-town paper. Therefore, the findings of this study support the utility of integrating the traditional source effects perspective with that of the HME, suggesting the need for further inquiry.
The scales measuring bias against the home team and bias against the rival team represent traditional measures of HME specific to one side of the issue or to one partisan group. The third measure, comparative bias in favor of rival team, represents a forced, final judgment of the story, requiring participants to judge which group seems more maligned by the article. As previous HME research has noted,40 some stories will be seen by all receivers (even neutral) as somewhat slanted. Eventually, even if both sides seem to be presented unfavorably one must arrive at a relative judgment on which side the story is more hostile toward; this comparative measure may be the most appropriate measure of the final assessment of a story.
While the test of H3 indicated that HME might be specific to the newspaper in which the story was printed, the tests of how such judgments were made are less clear. Different sources of the story were not associated with significant differences in most of the measures that assessed the manner in which the content of the story was processed. However, retrospective estimates of the percentage of negative statements about the home team in the article tended to be lower (at p < .10) in the home-town paper condition than in the other two paper conditions. This finding does provide limited evidence of attitude-based processing of content suggested by some researchers as a possible cause of the HME. In a previous study, higher retrospective estimates of negative references to a group were found to be associated with prior beliefs about media bias.41 The authors suggested that participants may have employed the following implicit reasoning strategy: "If the media favor the opposite side...then this newscast has to contain more hostile than congenial references to my side."42 The findings reported here suggest a similar process: participants in the home-town-paper condition might have expected from their local paper less bias against the home team and, therefore, retrospectively assumed fewer negative references to their team. However, this assumption cannot be substantiated due to lack of correlation between the prior beliefs about media bias measure (perhaps too broadly defined, as described below) and perceptions of story bias.
While variation of the news source resulted in differences in recall and processing of story content for just one variable (retrospective estimates of percentage of negative references to the home team), the judgment-heuristic explanation for the HME was completely unsupported: neither partisanship nor previous beliefs about media bias were correlated with perceptions of hostility in any of the three conditions. Three explanations may account for the lack of evidence for the heuristic path to hostile media perceptions. First, while previous studies found small if any effects with less involved partisans or those lacking a strong identification with the group being investigated,43 the current study tapped the other end of the spectrum. The mean fanship score was 8.07 (s.d. = 2.54), and the median was 9.33. The extremely high levels of partisanship might have accounted for the lack of a relationship between partisanship and perceptions of bias.
Second, the media bias measure asked participants about general bias in media, rather than about specific bias toward their group. Beliefs about general media bias may not carry over to expectations of bias for narrow issues such as the one investigated here. Previous measures of bias have been specific to the group or issue at hand.44 While it seems implausible that beliefs about general media bias against a particular school football program would be prevalent, future studies might investigate such a possibility.
Finally, because only a subsample of participants completed the media bias pretest the number of cases included in the relevant analyses was small and the power of such analyses was limited. Because direct, strong paths from partisanship and prior beliefs of media bias have been identified in previous HME research, future studies would benefit from consideration of these variables. It should be noted that this study did not attempt to indicate an either/or answer to the research questions regarding mechanism. It is certainly possible that both mechanisms could be at work: heuristic paths to judgments do not necessarily preclude more central or systematic paths.45
The present study further illuminates the relationship between news source, prior attitudes about media bias, and the HME. If further studies support the findings herein, media outlets would be faced with a challenging proposition: news consumers may expect more congenial coverage of some issues from local media. Sports news certainly seems to be an area in which this relationship might be manifest. One could also conceive of the same expectation of congenial coverage for negative, local events that gain national attention (e.g., the confederate flag issue in South Carolina, voting irregularities in Florida). If news consumers expect congenial coverage from local media, the journalistic mission of providing objective or balanced news coverage may be challenged. However, it is also possible that news consumer expectations of congeniality from local media will actually influence processing of story content, resulting in favorable perceptions of such media. Whatever the case, additional investigations of how the proposed mechanisms associated with the HME might vary depending upon story source seem warranted.
J&MC Quarterly Vol. 80, No. 2 Summer 2003 265-281 (C)2003 AEJMC
1. Wayne Wanta and Yu-Wei Hu, "The Effects of Credibility, Reliance, and Exposure on Media Agenda Setting: A Path Analysis Model," Journalism Quarterly 71 (spring 1994): 90-98.
2. Pew Research Center For the People and the Press, Terror Coverage Boosts News Media's Images, 28 November 2001, http://people-press.org/reports/display.php3?ReportID= 143.
3. Cecilie Gaziano and Kristen McGrath, "Measuring the Concept of Credibility" Journalism Quarterly 63 (autumn 1986): 451-62.
4. For instance: Cindy T. Christen, Parthana Kannaovakun, and Albert C. Gunther, "Hostile Media Perceptions: Partisan Assessments of Press and Public During the 1997 United Parcel Service Strike," Political Communication 19 (October 2002): 423-36; Roger Giner-Sorolla and Shelly Chaiken, "The Causes of Hostile Media Judgments," Journal of Experimental Social Psychology 30 (March 1994): 165-80; Albert C. Gunther, "Biased Press or Biased Public? Attitudes Toward Media Coverage of Social Groups," Public Opinion Quarterly 56 (summer 1992): 147-67; Albert C. Gunther and Cindy T. Christen, "Projection or Persuasive Press? Contrary Effects of Personal Opinion and Perceived News Coverage on Estimates of Public Opinion," Journal of Communication 52 (spring 2002): 177-95; Albert C. Gunther, Cindy T. Christen, Janice L. Liebhart, and Stella C. Chia, "Congenial Public, Contrary Press, and Biased Estimates of the Climate of Opinion," Public Opinion Quarterly 65 (fall 2001): 295-320; Richard M. Perloff, "Ego-Involvement and the Third Person Effect of Televised News Coverage," Communication Research 16 (summer 1989): 236-62; Robert Vallone, Lee Ross, and Mark Lepper, "The Hostile Media Phenomenon: Biased Perception of Media Bias in Coverage of the Beirut Massacre," Journal of Personality and Social Psychology 49 (fall 1985): 577-85.
5. Gunther et al., "Congenial Public."
6. Charles G. Lord, Lee Ross, and Mark R. Lepper, "Biased Assimilation and Attitude Polarization: The Effects of Prior Theories on Subsequently Considered Evidence," Journal of Personality and Social Psychology 37 (May 1979): 2098-2109.
7. Albert H. Hastorf and Hadley Cantril, "They Saw a Game: A Case Study," Journal of Personality and Social Psychology 49 (1954): 129-34.
8. See Joseph T. Klapper, The Effects of Mass Communication (Glencoe, IL: Pree Press, 1960); Richard Nisbett and Lee Ross, Human Inference: Strategies and Shortcomings of Social Judgment (Englewood Cliffs, NJ: Prentice Hall, 1980); Neil Vidmar and Milton Rokeach, "Archie Bunker's Bigotry: A Study in Selective Perception and Exposure," Journal of Communication 24 (winter 1974): 36-47.
9. Vallone, Ross, and Lepper, "The Hostile Media Phenomenon," 579.
10. Perloff, "Ego-Involvement."
11. Vallone, Ross, and Lepper, "The Hostile Media Phenomenon."
12. Vallone, Ross, and Lepper, "The Hostile Media Phenomenon."
13. Giner-Sorolla and Chaiken, "The Causes of Hostile Media Judgments"; Perloff, "Ego-Involvement."
14. Gunther, "Biased Press."
15. Gunther et al., "Congenial Public."
16. Gunther et al., "Congenial Public."
17. For more information: Shelly Chaiken, "Heuristic Versus Systematic Information Processing and the Use of Source Versus Message Cues in Persuasion," Journal of Personality and Social Psychology 39 (November 1980): 752-56; Shelly Chaiken and Alice Eagly, "Communication Modality as a Determinant of Persuasion: The Role of Communicator Salience," Journal of Personality and Social Psychology 45 (August 1983): 241-256; Richard E. Petty and John T. Cacioppo, Communication and Persuasion: Central and Peripheral Routes to Attitude Change (NY: Springer-Verlag, 1986).
18. Vallone, Ross, and Lepper, "The Hostile Media Phenomenon."
19. Giner-Sorolla and Chaiken, "The Causes of Hostile Media Judgments"; Vallone, Ross, and Lepper, "The Hostile Media Phenomenon."
20. Gunther, "Biased Press."
21. See Cecilie Gaziano, "How Credible is the Credibility Crisis?" Journalism Quarterly 65 (summer 1988): 267-78; Gaziano and McGrath, "Measuring the Concept of Credibility"; Philip Meyer, "Defining and Measuring Credibility of Newspapers: Developing an Index," Journalism Quarterly 65 (fall 1988): 567-74.
22. Research on source effects has indicated that the effectiveness of communication is a function of the audience's attitude toward the source of the message; for more see Carl I. Hovland, Irving L. Janis, and Harold H. Kelley, Communication and Persuasion (New Haven, CT: Yale University Press 1953); and Carl I. Hovland and Walter Weiss, "The Influence of Source Credibility on Communication Effectiveness," Public Opinion Quarterly 15 (winter 1951): 635-50. Although early conceptualizations of source effects have been challenged for the assumption that credibility is inherent to the source, rather than an individual's or group's situationally bound perception of a source (cf. David K. Berlo, James B. Lemert, and Robert J. Mertz, "Dimensions for Evaluating the Acceptability of Message Sources," Public Opinion Quarterly 33 [winter 1969]: 563-76; Jesse G. Delia, "A Constructivist Analysis of the Concept of Credibility," Quarterly journal of Speech 62 [December 1976]: 361-75; Gunther, "Biased Press"; and James C. McCroskey, "Scales for the Measurement of Ethos," Speech Monographs 33 [March 1966]: 65-72), decades of source effects research embracing the latter conceptualization have reliably demonstrated that audiences judge a message more favorably when it is delivered by a source perceived as having a relatively high, rather than low, level of credibility (for more see Richard M. Perloff, The Dynamics of Persuasion [Hillsdale, NJ: Erlbaum, 1993]; and Elizabeth Wilson and Daniel Sherrell, "Source Effects in Communication and Persuasion Research: A Meta-Analysis of Effect Size," Journal of the Academy of Marketing Science 21 [summer 1993]: 101-112).
23. Gunther, "Biased Press."
24. Gunther, "Biased Press."
25. Initial conceptualizations of the hostile media effect involved a comparison of a partisan group's perception of story bias to a non-partisan group's perception. However, more recent conceptualizations, such as that in the study by Gunther et al. ("Congenial Public") suggest that because even neutral parties may sometimes consider a story slanted, a relative comparison of bias is more appropriate. The relative hostile media effect, therefore, is defined as the case when two groups on two sides of an issue review the same story but each thinks it is more biased against its own side. As the current study examines only one group of partisans, the original hostile media effect conceptualization is retained.
26. Gunther et al., "Congenial Public."
27. For a comprehensive summary, see Daniel L. Wann, Merrill J. Melnick, Gordon W. Russell, and Dale G. Pease, Sports Pans: The Psychological and Social Impact of Spectators (NY: Routledge, 2001).
28. See Robert B. Cialdini, Richard J. Borden, Avril Thorne, Marcus Randall Walker, Stephen Freeman, and Lloyd Reynolds Sloan, "Basking in Reflected Glory: Three (Football) Field Studies, Journal of Personality and Social Psychology 34 (September 1976) : 366-75; Robert B. Cialdini and Kenneth D. Richardson, "Two Indirect Tactics of Image Management: Basking and Blasting," Journal of Personality and Social Psychology 39 (September 1980): 406-415.
29. See C. R. Snyder, Raymond L. Higgins, and Rita J. Stucky, Excuses: Masquerades in Search of Grace (NY: Wiley-Interscience, 1983); C. R. Snyder, Mary Anne Lassegard, and Carol E. Ford, "Distancing After Group Success and Failure: Basking in Reflected Glory and Cutting Off Reflected Failure," Journal of Personality and Social Psychology 51 (August 1986): 382-88.
30. Edward R. Hirt, Dolf Zillmann, Grant A. Erickson, and Chris Kennedy, "Costs and Benefits of Allegiance: Changes in Fans' Self-Ascribed Competencies after Team Victory Versus Defeat," Journal of Personality and Social Psychology 63 (November 1992): 724-38.
31. Hastorf and Cantril, "They Saw a Game."
32. Karla Schweitzer, Dolf Zillmann, James B. Weaver, and Elizabeth S. Luttrell, "Perception of Threatening Events in the Emotional Aftermath of a Televised College Football Game," Journal of Broadcasting & Electronic Media 36 (winter 1992): 75-82.
33. For more information on value-relevant involvement, see Blair T. Johnson and Alice H. Eagly, "Effects of Involvement on Persuasion: A Meta-Analysis," Psychological Bulletin 106 (September 1989): 290-314.
34. For a comprehensive summary of attribution theory, see Harold Kelley, "The Process of Causal Attribution," American Psychologist 28 (February 1973): 107-128.
35. Alice Eagly, Shelly Chaiken, and Wendy Wood, "An attribution analysis of persuasion," in New Directions in Attribution Research, vol. 3, ed. John H. Harvey, William Ickes, and Robert F. Kidd (Hillsdale, NJ: Erlbaum, 1981), 37-62. See also Alice Eagly, Wendy Wood, and Shelly Chaiken, "Causal Inferences about Communicators and their Effect on Opinion Change," Journal of Personality and Social Psychology 36 (April 1978): 424-35.
36. Vallone, Ross, and Lepper, "The Hostile Media Phenomenon," 579.
37. Christen, Kannaovakun, and Gunther, "Hostile Media Perceptions"; Giner-Sorolla and Chaiken, "The Causes of Hostile Media Judgments"; Vallone, Ross, and Lepper, "The Hostile Media Phenomenon."
38. The questionnaire indicated that "roughly 50% of the statements actually appeared in the article."
39. For example, see Gaziano and McGrath, "Measuring the Concept of Credibility" and Meyer, "Defining and Measuring Credibility."
40. Gunther et al., "Congenial Public."
41. Giner-Sorolla and Chaiken, "The Causes of Hostile Media Judgments."
42. Giner-Sorolla and Chaiken, "The Causes of Hostile Media Judgments," 177.
43. Giner-Sorolla and Chaiken, "The Causes of Hostile Media Judgments"; Vallone, Ross, and Lepper, "The Hostile Media Phenomenon."
44. Giner-Sorolla and Chaiken, "The Causes of Hostile Media Judgments."
45. For more on this, see Alice Eagly and Shelly Chaiken, The Psychology of Attitudes (Fort Worth, TX: Harcourt Brace, 1993); and Eric P. Thompson, Arie W. Kruglanski, and Scott Spiegel, "The Unimodal and Attitudes," in Why We Evaluate: Functions of Attitudes, ed. Gregory R. Maio and James M. Olson (Mahwah, NJ: 2000, Erlbaum), 59-96.
Laura M. Arpan and Arthur A. Raney teach at Florida State University in the Department of Communication.…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: An Experimental Investigation of News Source and the Hostile Media Effect. Contributors: Arpan, Laura M. - Author, Raney, Arthur A. - Author. Journal title: Journalism and Mass Communication Quarterly. Volume: 80. Issue: 2 Publication date: Summer 2003. Page number: 265. © Association for Education in Journalism and Mass Communication Winter 2007. Provided by ProQuest LLC. All Rights Reserved.