Content-Type: text/html

This paper was presented at the Association for Education in Journalism and
Mass Communication in San Antonio, Texas August 2005.
         If you have questions about this paper, please contact the author
directly. If you have questions about the archives, email
rakyat [ at ] eparker.org. For an explanation of the subject line, 
send email to
<a href="/cgi-bin/wa?LOGON=A3%3Dind0602A%26L%3DAEJMC%26E%3D8bit%26P%3D8068903%26B%3D--%26T%3Dtext%252Fplain%3B%2520charset%3Diso-8859-1" >[log in to unmask]</a> with just the four words, "get help info aejmc," in the
body (drop the "").

(Feb 2006)
Thank you.
Elliott Parker
====================================================================

Ideology and Source Credibility:
Partisan Perception Bias in Believability of CNN, Fox News and PBS
By
Zengjun Peng

Doctoral Student
Missouri School of Journalism

1000 Hospital Dr.  Apt. 1J
Columbia, MO65201
Phone: (573) 771-0077
Email: <a href="/cgi-bin/wa?LOGON=A3%3Dind0602A%26L%3DAEJMC%26E%3D8bit%26P%3D8068903%26B%3D--%26T%3Dtext%252Fplain%3B%2520charset%3Diso-8859-1" >[log in to unmask]</a>

Submission for Leslie J. Moeller Award Competition
Mass Communication and Society Division
AEJMC Annual Convention, San Antonio, Texas, 2005
Ideology and Source Credibility:
Partisan Perception Bias in Believability of CNN, Fox News and PBS

This paper examines the relationship between partisan ideology and 
perception of source believability within the framework of hostile 
media effect.  Results show that partisan ideology significantly 
influenced people's perceived believability of three news outlets of 
CNN, Fox News and PBS (News Hour with Jim Lehrer).  Liberals are more 
likely to rate CNN as believable while conservatives tend to endorse 
Fox News. Partisan ideology, however, does not make a difference in 
the evaluation of PBS.
Ideology and Source Credibility:
Partisan Perception Bias in Believability of CNN, Fox News and PBS
Perhaps one of the most unsettling issues in journalism is the 
accusation of media bias, and together with that, the growing public 
distrust.  Although millions of Americans still rely on the news 
media as their major source of information (Pew, 2003), the public 
seems to simultaneously have a growing distrust on the news media 
(Pew, 2002, 2003). For instance, when the Gallup Poll asked Americans 
to rate the honesty and ethics of people in various fields, 
journalists were rated ahead of only car dealers and insurance 
salesmen (Gallup, 2000). While the news institutions and journalists 
who work for them claim objectivity as one of their highest 
standards, studies have shown that more and more Americans are 
skeptical of the news media as a "fair," "trustworthy," and impartial 
source of political information, and the news media have been 
perceived as politically and ideologically biased (ASNE, 1999; 
Cappella & Jamieson, 1997; Pew Center, 2002, 2003).
Conventional source credibility studies tend to conceptualize media 
credibility along two dimensions: source credibility and content 
believability. Overlapping but distinctive as the two concepts are, 
source credibility often seems to be more inclusive, embracing 
measurement of both trustworthiness of the source and realism of the 
content. Thus, source credibility has been studied in relation with a 
number of media institutional characters, and media content, 
including the reputation of the news media, the professionalism of 
the journalists, objectiveness of the news stories, etc. However, 
less attention has been paid to the perceptive quality of source 
credibility. Often than not, source credibility is perceptive rather 
than objective, and such perception is subject to a number of 
personal factors, most prominently political and ideological 
dispositions (Chaiken, Liberman, & Eagly, 1989; Christen, 
Kannaovakun, & Gunther, 2002; Higgins & Bargth, 1987; Vallone, Ross, 
& Lepper, 1985).
This study adopts this perceptive line of thinking and examines the 
relationship between ideological factors and perception of news media 
credibility within the framework of hostile media effect (Vallone, 
Ross, & Lepper 1985). Using secondary data from the Pew center media 
believability survey (Pew, 2002), this study tests how individual 
ideology influences perceived believability of three news networks: 
CNN, Fox News and PBS (News Hour with Jim Lehrer).
  The hostile media effect was first identified and tested in a 
classic study by  Vallone and colleagues (1985). It holds that 
perception of media bias is in large part a result of the perceivers' 
ideological and political biases rather than the actual bias of news 
content. Partisans may still see biases even when the news coverage 
is balanced and objective. Although the original hostile media effect 
study dealt with the concept of media bias, not media credibility, it 
is argued that media bias is an integral part of media credibility, 
and it is plausible to extend the basic theoretical proposition of 
the original hostile media effect to the study of source credibility.
	The original and later replications of the hostile media effect 
study (Giner-Sorolla & Chaiken, 1994;  Gunther, 1992; Perlof, 1989) 
basically assumed a neutral, anonymous news message in their inquiry, 
without giving adequate considerations to other conditioning factors 
such as source credibility in triggering, enhancing or inhibiting the 
hostile media effect. Although several studies (e.g. Giner-Sorolla & 
Chaiken, 1994) indicated the importance of identifying prior belief 
in media bias and included such factor in studying hostile media 
perception, they did not explain what contributed to the formation 
and direction of prior beliefs in news media bias. In addition, they 
overlooked the issue that audience perception of news media bias can 
be both general (the new media are biased as a whole) and specific 
(certain news media outlets are perceived as liberal while others as 
conservative), and such perceptions are often dependent on the 
relative political and ideological bias of the viewers (A liberal may 
perceive certain news outlet as conservative, while a conservative 
may credit the same news outlet as fair). It is important to note 
that the viewers are normally exposed to reputable news organizations 
rather than anonymous sources on a daily basis, and the viewers have 
their own perceptions of each news outlet. Thus, it is important to 
examine how individual factors influenced the perception of bias and 
credibility of different news outlets. Such inquiry may not only 
offer evidence for the presumptive source impact on hostile media 
perception but also extend the boundary of the hostile media theory 
as a whole.
Literature Review
	Although the seminal hostile media effect study was done 20 years 
ago, only a couple of follow-up studies have been conducted. Due to 
the scarcity in previous literature in applying hostile media theory 
in studying the relationship between political ideology and source 
credibility, greater efforts will be expended on discussing the basic 
theoretical propositions of the hostile media effect, and how these 
can be connected with the study of source credibility. It also should 
be pointed out that previous studies in hostile media effect dealt 
mainly with the concept of media bias rather than media credibility 
per se. Although the two are closely connected, and many previous 
inquires did not explicitly make the distinction (Gunter, 1992), they 
are not identical. Perception of bias does not automatically mean 
perception of media credibility. For instance, viewers may still 
perceive a news story as credible even when they question its 
objectivity (Peng & Thorson, 2004). With that precaution being taken, 
however, this study follows the conventional line of not making 
arduous endeavor to differentiate the two, because the major goal of 
this study is to examine the underlying relationship between partisan 
ideological factors and perception on media performance.
The Hostile Media Effect
In studying media bias, two approaches have been conventionally 
adopted. One focuses on the production side, often by means of 
surveying the political and ideological profiles of the journalists 
(e.g. Weaver & Wilhoit, 1996; Weaver, 1998), their daily 
newsgathering routines (Gans, 1988; White, 1950), and the social and 
political contexts in which news gathering and production are 
performed (Shoemaker & Reese, 1996). The other common approach is 
bias identification through content analysis. The first line of 
inquiry, although may provide important information on who 
journalists are and how they do their work, it is difficult to 
establish causal connections between journalists=s personal 
preferences and the bias of the news content. Content analyses, on 
the other hand, by systematically analyzing the end products, 
produced important data on multiple dimensions of the nature and 
characteristics of the news content. However, content analyses are 
often subject to a number of possible biases such as sampling, coding 
scheme and result interpretation. Although a number of studies have 
combined the two to look at the correlations (Watts, Domke, Shah, & 
Fan, 1999) causation was hard to establish, as often the case in 
survey research.
Recently, more studies have been using the theoretical propositions 
in social cognition and social judgment theory to examine the 
perceptive aspects of media bias. The basic contention is that media 
bias is subject to distortion of non-media factors such as partisan 
ideological bias, as demonstrated in the seminal hostile media effect 
by Vallone, et al (1985).
Prior to the study of perception of media bias, the issue of 
distorted evaluations and perception of social evidence by 
preconceived theories and beliefs have been documented in a great 
number of studies in social psychology (Allport, 1954; Bruner, 1957; 
Chapman & Chapman, 1967; Hastorf & Cantril, 1954; Hamilton, 1979; 
Nisbett & Ross, 1980; Lord, Ross, & Lepper, 1979; Ross & Lepper, 
1980). The classic study by Hastorf and Cantril (1954) recorded the 
influence of partisan bias on evaluation of a sport game. Students 
from Princeton and Dartmouth were asked to view a film of a rough 
game between the two college teams. Interestingly, Princeton students 
"saw" the Dartmouth team make over twice as many rule infractions as 
were seen by Dartmouth students. Hastorf and Cantril interpreted 
these results overall as indicating that, when encountering a mix of 
occurrences as complex as a football game, people experience 
primarily those events that fulfill a familiar pattern and have 
personal significance.
Similar phenomenon was documented in a later study by Lord, Ross and 
Lepper (1979). In that study, the researchers asked about advocates 
and opponents of capital punishment to review an identical pair of 
studies that provided mixed results on deterrent efficacy of such 
punishment. They found that the exposure to a common sample of 
inconclusive evidence produced no moderation or convergence of views 
on the part of the partisans. On the contrary, each group readily 
assimilated or accepted at face value the evidence that seemed to 
support its position, but subjected to critical scrutiny the evidence 
that threatened or undermined its position. This clearly demonstrated 
that issue positions, in this case, advocates and opponents of 
capital punishment, did influence their perception and judgment of 
the same evidences.
	The first study to explicitly identify the hostile media effect was 
done by Vallone, et al (1985). Vallone and colleagues recruited 
Stanford University students who identified themselves as generally 
pro-Israel, pro-Arab, and mixed neutral feeling to view news coverage 
of the Beirut massacre. After viewing identical samples of major 
networks  television coverage, both pro-Israel and pro-Arab partisans 
rated these programs, and those responsible for them, as being biased 
against their side.
Following Vallone, et al's (1985) seminal study, a number of studies 
have used both survey and experimental methods to examine different 
aspects of the hostile media effect, including two studies 
replicating Vallone, et al in using Palestine-Israeli conflict as the 
context (Giner-Sorolla & Chaiken, 1994; Perloff, 1989). Beginning in 
the late 1990s, additional field experiments on hostile media effects 
were conducted on some other issues such as perception on the 1997 
UPS strike (Christen, Kannaovakun, & Gunther, 2002), the controversy 
over lab research using primates (Gunther, Christen, Liebhart, & 
Chia, 2001), and presidential performance (D'Alessio, 2003). A number 
of survey studies, some of them based on national probability samples 
(Dalton, Beck, & Huckfeldt, 1998; Gunther, 1992; Gunther & Christen, 
2002) also showed results consistent with findings of Vallone, et al (1985).


News Source and Hostile Media Effect
Vallone, et al (1985) and most later hostile media studies did not 
include source as an important factor for consideration. In these 
studies, a neutral news source was assumed in the stimuli. However, 
source can greatly influence the perception process. For instance, 
Rouner, Slater, and Buddenbaum (1999) demonstrated how readers 
consider the source in evaluating whether a statement is biased. They 
provided readers with short biographies of speakers written to 
suggest the speaker would be biased or unbiased on the issue. They 
found that, as predicted, readers evaluated the speakers statements 
based on what they had been told about the source, rather than the 
actual statement content. Related research also indicated the 
importance of source in perception of news credibility (Gunther, 
2002: Gaziano, 1988; Gaziano & McGrath; Meyer, 1988). Source effects 
also bear strong effect on the effectiveness of communication found 
in earlier researches (Hovland, Janis, & Kelly, 1953; Hovland & 
Weiss, 1951). Later studies 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 
(Perloff, 1993; Wilson & Sherrell, 1993).
Source effect in hostile media phenomenon was first tested in a 
recent study by Arpan and Raney (2004). The study examined the 
interaction among different news sources, and the hostile media 
effect in sports news. The news source manipulation was achieved by 
altering the location of the newspaper that published the story. 
Participants read a balanced story about their hometown college 
football team in one of three newspapers: the hometown, the cross 
state rival university town, or a neutral town paper. As expected, 
the participants=s perception of bias was significantly influenced by 
the perceived news source.
Partisanship and News Source Credibility
News source credibility has been traditionally examined as an 
inherent characteristic of the news institution, dependent or 
independent of the news content. A large part of the research focused 
on the definition and measurement of credibility (Gaziano & McGrath, 
1986). Source credibility has been conceptualized as the degree to 
which an individual perceives the media source portrays the real 
world truthfully (Austin & Dong, 1995). Gunther and colleagues 
(Gunther & Lasorsa, 1986; Gunther, 1988) defined the concept of 
credibility on two dimensions: media expertise in covering a topic 
and perceptions about the media bias in covering the topic. All 
these, although different in emphases, point to the perceptive 
dimensions that audience exploited to make judgment.
Although the hostile media effect studies did not explicitly point 
out how partisan factors influence perception of media credibility, 
the discussion on source effect in previous hostile media effect 
research did indicate that similar relationship existed. For 
instance, studies show that perception of news media credibility may 
be less a matter of news content than cognitive process of the 
audiences. (Christen, et al, 2002; Giner-Sorolla & Chaiken, 1994; 
Gunther, 1992). News stories intended as balanced and objective by 
the news producers may be perceived as biased, and therefore lacking 
in credibility by the audience. Along this line of argument, if the 
partisan bias played a critical role in perception of media bias, 
there is possible too that such mechanism may be present in 
perception of credibility of news outlets.


Ideological Spectrum of CNN, Fox News and PBS
There are many plausible factors as why partisanship may impact their 
perception of news media credibility. One line of argument is the 
perceived media bias, which impacts people's  perception of news 
media credibility, as indicated by the hostile media effect theory. 
As shown in the study of Arpan and Raney (2004) in sports news 
coverage, audience perception was greatly influenced by the location 
of the news sourceBpaper of hometown, of rival town, and of a neutral 
town.  Apparently, the presumed media bias in news sources influenced 
the judgment process. Similarly, as in this study, the three news 
networks may be perceived differently on the ideological slant by the 
audience. Although placement of these three networks on the 
ideological continuum may be a matter of debate and the branding of 
CNN as the liberal, Fox as conservative are more based on popular 
accusations and perceptions rather than hard empirical evidence, it 
seems fair to say that these classifications do exist as far as 
public perception is concerned.  Generally speaking, Fox news has 
been identified with a conservative slant, therefore, was used here 
as a representative on the right end of the ideological continuum of 
the news media; similarly, CNN is on the left end of the continuum, 
and PBS as somewhere in between. Although in reality, such as 
categorization or placement of the news channel might not be accurate 
and difficult to justify, this study argues that what matters here is 
not whether such placement is accurate, but the popular perceived 
ideological and political bias of the news media. After all, in this 
study, it is the relative perceptions of the viewers rather than the 
actual bias of the news media are measured and tested.
The study of source credibility is important because the audience, 
for the most part, is only receiving partial information, and seldom 
has a baseline for evaluating the information they get. And rarely, 
if not at all, will average viewers bother to expose themselves to 
news coverage of different sources on the same event in order to 
balance out any possible bias. To them, perception on the specific 
news outlets can be vital in evaluating the news content.
Based on the above discussion, it is reasonable to expect hostile 
media effect on source credibility as indicated by the original 
hostile media hypothesis on perceived media bias. Specifically, 
partisan, ideological factors will result in bias perception of 
specific news sources. Rather than perceiving a hostile media bias 
across media (the news media are generally biased either liberally or 
conservatively), subjects will differentiate their perception towards 
specific news media outlets. For instance, conservatives will be more 
likely to perceive a hostile bias in CNN than Fox News; or from a 
relative perspective, both liberals and conservatives may perceive a 
hostile media bias in CNN, but conservatives will be more likely to 
perceive a stronger hostile effect in CNN than liberals.
The data used in this study was from a media believability survey 
done in the year 2002 by Pew Center. Independent variables chosen for 
this study include demographics (age, gender, education, income), 
Party ID (Republican, Independent, or Democrats), ideology 
(conservative or liberal), and exposure to national TV news. The 
dependent variable is perceived believability of CNN, PBS, and Fox News.
It is expected that liberals would perceive CNN as more believable 
and conservatives would perceive Fox News as more believable. 
However, there will be no significant differences between the two 
camps on their perception of PBS news. Specifically:
H1. There will be a significant positive relationship between how 
liberal a respondent is and the perceived believability of CNN.
H2. There will be a significant negative relationship between how 
liberal a respondent is and the perceived believability of Fox News.
H3. There will be no significant relationship between how liberal a 
person is and the perceived believability of PBS. In other words, 
there will be no significant differences between liberals and 
conservatives in their perception of the believability of PBS.
Methods
The Pew Center is an independent opinion research group that studies 
attitudes towards the press, politics and public policy issues. It 
sponsors regular national surveys that measure public attentiveness 
to major news stories and charts trends in values and fundamental 
political and social attitudes. Its data were open for public access 
and have been used for analysis in journalism, communication, 
political sciences and many other areas of academic research.
The 2002 Media Believability Survey are based on telephone interviews 
conducted under the direction of Princeton Survey Research Associates 
among a nationwide sample of 1,005 adults, 18 years of age or older, 
during the period May 6-16.
The dataset contains hundreds of variables measuring a wide range of 
demographic and attitudinal measures. For this study, only variables 
pertinent to the inquiry were selected, including demographics, party 
ID, political views, exposure to national TV news, and believability 
of specific news outlets.
Party ID was measured with two related questions. First, respondents 
were asked: In politics today, do you consider yourself a Republican, 
Democrat, or Independent? If in independent was chosen, a follow-up 
question was asked for the learning direction (Republican or Democrat).
Political view was operationalized in a similar way, with a question 
asking the respondents to rate their political view on a five-point 
scale ranging from 1 (very conservative) to 5 (very liberal).
National TV news exposure was operationalized by a four-point scale 
ranging from 1 (rarely) to 4 (regularly).
Believability was operationalized with a single question: how much do 
you think you can believe each of the following organization? 
Respondents were asked to rate on a 4-point scale ranging from 1 
(can't  believe) to 4 (believe the most)
Multiple regressions models with block entry were tested on CNN, Fox 
News, and PBS, with demographics and party ID as the first block, 
followed by TV news exposure, and political view.
Results
	Descriptive statistics (See Table 1) show that the sample generally 
represented the national population (United States Census, 
2000).  The sample had  slightly more females (male =46.3 percent, 
and female =53.7). In terms of party ID, there were roughly equal 
numbers of Republicans (31.6 percent) and Democrats (31.4 percent), 
with the rest as Independents. They were generally regular network 
news consumers (mean=2.97 in the range of 1-4). In terms of political 
view, there was a roughly normal distribution, with a mean of 2.77, 
and a standard deviation of .97 in the range of 1-5. In terms of 
believability (ranging from 1 to 4), CNN (mean=3.03) led the three, 
followed by PBS (mean=2.87) and Fox News (mean=2.81). However, the 
differences were minimal.

Insert Table 1 about Here

	Table 2 presents the zero-order correlations between believability 
and the independent variables. For CNN, sex, TV news exposure and 
political views are significantly associated with perception of 
believability in the positive direction. However, age is negatively 
associated with believability. This indicates that younger, female 
viewers with liberal political views tend to perceive CNN as more 
credible. Education, income and party ID have no significant 
correlations with believability. The results for Fox News viewers 
turn out to be contrasting: all the seven variables are significantly 
associated with perceived believability. However, different from CNN, 
age is in the positive direction, indicating older viewers tend to 
perceived Fox as more believable. Education level, which is not 
significant in CNN, becomes significant, negatively correlating with 
perception of believability, which can be interpreted that the more 
educated, the less likely a person will rate Fox News as believable. 
Party ID is also negatively associated with believability of Fox 
News. In other words, the Democrats are less likely to rate Fox news 
as believable. Not surprisingly, political view is negatively 
correlated with perception of believability of Fox News: 
conservatives are more likely to perceive Fox News as believable. For 
PBS, age, sex and education are positively associated with perceived 
believability. Party ID, TV news exposure, and particularly political 
views, are not correlated with perceived believability.

Insert Table 2 about Here
	Table 3 shows the results of multiple regressions, with demographics 
(age, sex, education, income) and Party ID as the first block, 
national TV news exposure as the second, and finally political views. 
Among the demographics, the relationships found in zero-order 
correlations still hold, although the strength is limited. 
Interestingly, sex is a strong predictor (beta=.205, p<.01 for CNN; 
beta=.079, p<.01 for Fox, and beta= .152, p<.01 for PBS); Age is a 
negative predictor for CNN (beta=-.162, p<.05), but a positive one 
for PBS (beta=.143, p<.01). Education is a negative predictor for CNN 
(beta=-.054, p<.01) and Fox (-.098, p<.05), but a positive one for 
PBS (beta=.183, p<.01). Income is a negative predictor for Fox News 
(beta=-.097, p<.05). While Party ID is not significant for CNN and 
PBS, it is negatively predicting Fox (beta=-.092, p<.05). National TV 
news exposure is a strong predictor for both CNN (beta=.166p,.p<.01) 
and Fox News (beta=.234, p<.01), but not for PBS. TV news exposure, 
not surprisingly, strongly predicts believability of  CNN (beta=.166, 
p<.01) and Fox News (beta=.234, p<.01), but not PBS. Finally, 
political view is a significant predictor for both CNN (beta=.89, 
p<.01) and Fox (beta=-.174, p<.01) but in the opposite direction. 
Political view does not predict believability of PBS. The entire 
model explains about 10 percent variance for CNN, 12 percent for Fox news.

Insert Table 3 about Here

H1 predicts that there will be a significant positive relationship 
between how liberal a respondent is and the perceived believability 
of CNN. The results of the regression show that with all other 
variables controlled, political view positively predicts the 
believability of CNN. In other words, the more liberal a person is, 
the more the person will rate CNN as believable. Therefore, H1 is accepted.
H2 predicts an opposite direction for Fox News. Results show a 
significant confirmation of the prediction. With all other variables 
controlled, political view negatively predicts the perceived 
believability of Fox.  In other words, the more liberal a person is, 
the less likely the person will rate Fox as believability. Put it 
another way, conservatives will be more likely to perceive Fox news 
as believable. Therefore, H2 is accepted.
H3 states that there will be no significant relationship between how 
liberal a person is and the believability of PBS. In other words, 
there will be no significant difference between liberals and 
conservatives in their perception of the believability of PBS. The 
results show that political view is not a significant predictor for 
believability of PBS. Therefore, H3 is also accepted.

Discussion
	The key issue under investigation in this study is the relationship 
between one's ideology and perceived believability of individual news 
outlets that are perceived in relative ideological distance from 
one's own. The results in this study show that with all the effects 
of demographic variables and media exposure controlled for, political 
view significantly predicts believability perception. As expected, 
liberals tend to perceive CNN as more believable, and conservatives 
give more credits to Fox News, fitting the line of the hostile media 
effect. It's interesting to find that political view makes no 
difference when it comes to the evaluation of PBS.
	In addition to the significant key finding discussed above, it's 
also worth mentioning that among other predictors in the multiple 
regression models, TV news exposure turned out to be the strongest 
positive predictor for believability of both CNN and Fox News. It's 
not surprising in light of the numerous findings in mass 
communication literature that exposure and favorability are in a 
reciprocal relationship. Exposure can cultivate and enhance cognitive 
and affective favorability, and such favorability can be further 
translated into more exposure. It's also interesting to note that 
party ID weakly predicts believability of Fox News, not CNN. In 
contrast, ideology is a much stronger predictor for both Fox News and 
CNN. This indicates that although the concepts of party ID and 
ideology can be overlapping in many ways, they are distinctive under 
many situations, where ideology rather than party ID is a more viable 
concept.
	News media outlets are not identical in terms of ideological slants, 
and the audience's  perception of individual news media outlets 
matters. If one perceives the news source as with his or her side, 
one tends to believe, and rate it as fairer and less biased; however, 
if the news source is perceived as hostile, the credibility of the 
news source is more likely to be negatively perceived.  If the media 
source is perceived as objective or non-partisan, then people's own 
political and ideological bias does not make much difference. 
Although the total variance explained by the model is moderate, the 
results are notable given the complexity of the perception issue.
	A basic theoretical assumption of the hostile media effect is that 
audience evaluation of news media bias and credibility is perceptive 
and influenced by a plethora of personal factors, most notably 
partisan bias. The results of this study provide further 
evidence.  However, the original line of testing on the hostile media 
effect has been focused on the evaluation of actual news content 
rather than the news source. Most of the previous inquiry was testing 
how audiences' processing bias influenced assessment of bias of the 
news media coverage on specific issues. This study shows that the 
hostile media effect can extend to more general perception issues, 
such as news source, which may not be directly related to specific 
news content. Rather than an interaction between people's  own 
political ideological bias and the news media content bias, it seems 
here that people's own political bias interacts with their beliefs in 
the news source, which results in their final evaluation of the 
credibility of individual news outlets.
	It has to be emphasized here that the placement of the three 
networks along the ideological continuum is based more on relative 
popular perception rather than the actual news slant in reality, 
although studies did show Fox news tends to be more opinionated, 
pro-America, and pro-conservative in comparison with other cable news 
networks (State of the News Media, 2005). It is likely that average 
viewers were influenced by the general perception and identified 
themselves with respective news outlets. Such identification can 
serve as a reference for individual viewers to categorize news 
outlets as "us" or "them," and consequently bias their evaluation of 
the bias and credibility of the individual news outlets. This perhaps 
explains why PBS was not predicted by the audience's political views, 
as PBS might be perceived as a news outlet of neither "us" nor 
"them."  Similar situations have been found Arpan and Raney (2003) on 
source effect on sports news coverage. Findings in this study further 
demonstrate that partisan bias may be an important factor impacting 
people's perception of news media credibility. On all accounts, such 
perception can be crucial in light of the news media's role in social 
and political communication. Previous studies show that audience tend 
to have more exposure and give more attention when they perceive the 
news media as more credible (Wanta & Hu, 1994), and it can also 
condition the agenda-setting function of the news media (Iyengar & 
Kinder, 1987; Wanta, 1997).
	The findings in this study provide insights for a better 
understanding of the theoretical capability and methodological issues 
in hostile media effect research. Most previous hostile media effect 
studies relied on experimental methods, where groups of partisan 
extremity (e.g., strong pro-Israel vs. strong pro-Palestine, in 
Vallone, et al, 1985; strong pro-life vs. strong pro-choice in 
Giner-Sorolla & Chaiken, 1994) were asked to react to coverage on 
highly volatile, provocative issues. Although these studies offered 
valuable empirical evidence and theoretical explanations for the 
hostile media phenomenon, the restrictive conditions in these lab 
experiments reduced the generalizability of the results and the 
applicability of the theoretical dispositions. This study used survey 
data from a national sample, and identified and supported a strong 
effect of partisan bias on source credibility. This indicates that 
hostile media effect can go much beyond what had been defined in the 
original line of inquiry—partisan bias influenced perception of 
specific new stories. It can also influence people's perceptions on 
other dimensions related to news media performance such as source credibility.
	It should be noted that the nature of the secondary data limited the 
study's ability to empirically explore and test possible mechanism 
involved in the perception of source credibility. For instance, 
although a significant relationship was found between partisan 
ideological orientations and perceived credibility of individual news 
outlets, causality is hard to establish. Although it is plausible 
that audience perceived the individual news outlets as liberal, 
conservative, and fairly balanced, and such perceived news media bias 
influenced people's evaluation of news media credibility, as 
suggested in the original hostile media effect, such prior held 
beliefs of the audience in this study were assumed rather than 
empirically measured and tested. Future studies may use direct 
measurement on perceived news bias of each individual news outlets, 
and examine how these prior beliefs interact with ideological factors 
and influence perception of news source credibility.
References
AT. R
Chaiken, S., Liberman, A., & Eagly, A. H. (1989). Heuristic and 
systematic processing within and beyond the persuasion context. In J. 
S. Uleman & J. A. Bargh (Eds.), Unitended thought  (pp. 212-252). New 
York: Guilford.
Christen, C. T., Kannaovakun, P.,& Gunther, A. C. (2002). Hostile 
media perceptions: Partisan assessments of press and public during 
the 1997 UPS strike. Political Communication, 19, 423-436.
Dalton, R., Beck, P. and Hckfeldt, R. (1998). Partisan cues and the 
media: Information flows in the 1992 presidential election. American 
Political Science Review. 92, (1): 111-126.
D'Alessio, D. (2003). An experimental examination of readers' 
perceptions of media bias. Journalism and Mass Communication 
Quarterly, 80, 282-294.
Gallop (2000). Gallup honesty and ethics in professions report, November 2000.
(http://www.gallup.com/poll/findings/indhnsty_ethics2asp).
Gans, Herbert. 1979. Deciding what=s news: A study of CBS Evening 
News, NBC Nightly News, Newsweek and Time. NY: Vintage.
Gaziano, C., & McGrath, K. (1986). Measuring the concept of 
credibility. Journalism Quarterly, 63, 451-462.
Giner-Sorolla, R., & Chaiken, S. (1994). The causes of hostile media 
judgment. Journal of Experimental Social Psychology, 30, 165-180.
Graber, D. A. (1993). Mass media and American politics, 4th ed. 
Washington, D.C.: CQ Press.
Gunther, A. C. (1992). Biased press or biased public? Attitudes 
toward media coverage of social groups. Public Opinion Quarterly, 56, 147B167.
Gunther, A. C. (1998). The persuasive press inference: Effects of 
mass media on perceived public opinion. Communication Research, 25, 486B504.
Gunther, A. C., Christen, C. T., Liebhart, J.,& Chia, C.-Y. S. 
(2001). Congenial public, contrary press, and biased estimates of the 
climate of opinion. Public Opinion Quarterly, 65, 295-320.
Gunther, A. C.,& Christen, C. T. (2002). Projection or persuasive 
press? Contrary effects of personal opinion and perceived news 
coverage on estimates of public opinion. Journal of Communication, 
52, 177-195.
Gunther, A. C., & Schmitt, K. M. (2004). Mapping boundaries of the 
hostile media effect. Journal of Communication, 54, 55-70.
Hastorf, A. H., & Cantril, H. (1954). They saw a game: A case study, 
Journal of Abnormal and Social Psychology.
Higgins, E. T., & Bargh, J. (1987). Social cognition and social 
perception, Annual Review of Psychology, 38, 369-425.
Houston, D. A., & Fazio, R. H. (1989). Biased processing as a 
function of attitude accessibility: Making objective judgments 
subjectively. Social Cognition, 7, 51-66.
Hofstetter, C. R. (1976). Bias in the news: Network television 
coverage of the 1972 election campaign. Columbus, OH: Ohio State 
University Press.
Iyengar, S., & Kinder, D. R. (1987). News that matters. Chicago: 
University of Chicago Press.
Kahneman, D. (2003). A perspective on judgment and choice: Mapping 
bounded reality. American Psychologist, 58(9), 697-720.
Kahneman, D., Slovic, P., & Tversky, A. (1982). Judgment under 
uncertainty: Heuristics and biases. Cambridge: Cambridge University Press.
Kressel, N. J. (1987). Biased judgments of media bias: A case study 
of the Arab-Israeli dispute. Political Psychology, 8: 211-227.
Lippmann, W. (1922). Public Opinion. New York: Simon & Schuster.
Lord, C. G., Ross, L., & Lepper, M. (1979). Biased assimilation and 
attitude polarization: The effects of prior theories on subsequently 
considered evidence. Journal of Personality and Social Psychology, 
37, 2098-2109.
Mason, L., & Nass, C. (1989). How partisan and non-partisan readers 
perceive political foes and newspaper bias. Journalism Quarterly, 66, 
564-571, 778.
Paletz, David. 1999. The Media in American Politics: Contents and 
Consequences. NY: Longman.
Perloff, R. M. (1989). Ego-involvement and the third person effect of 
televised news coverage. Communication Research, 16(2), 236-262.
Pew (2002).  Media Believability Survey.  The Pew Center for the 
People and Press,
Pew (2003). War Attitude Tracking Survey. The Pew Center for the 
People and Press, .
Sherif, M., & Hovland, C. I. (1961). Social judgment: Assimilation 
and contrast effects in communication and attitude change. New Haven, 
CT: Yale University Press.
Sherif, C. W., & Sherif, M. (Eds.) (1967). Attitude, ego-involvement, 
and change. New York: Wiley.
Shoemaker, P., Reese, S.D. (1996). Mediating the message: theories of 
influences on mass media content. 2nd ed. New York: Longman.
The State of the News Media (2005). The state of the news media: An 
annual report on American journalism. Project in Excellence in Journalism.
Tuchman, G. (`1972). Objectivity as strategic ritual: an examination 
of newsmen=s notions of objectivity. American Journal of Sociology, 
77, 660-679.
Tuchman, G. (1978). Making news: a study in the construction of 
reality. New York: Free Press.
United States Census (2000). United States Census Demographics 
http://www.census.gov/Press-Release/www/2002/demoprofiles.html. 
Accessed March 12, 2005.
Vallone, R. P., Ross, L., & Lepper, M. R. (1985). The hostile media 
phenomenon: Biased perception and perceptions of media bias in 
coverage of the Beirut massacre. Journal of Personality and Social 
Psychology, 49(3), 577-585.
Wanta, W. (1997). How people learn about important issues: the public 
and the national agenda. Mashwah, NJ: Lawrence Erlbaum.
        Wanta, W. & Hu, Y.W. (1994).The effects of credibility, 
reliance and exposure on media    agenda-setting: a path analysis 
model, Journalism Quarterly,71,90-98.
Watts, D., Shah, & Fan (1999).  Elite Cues and Media Bias in 
Presidential Campaigns.@ Communication Research, 26: 144-175.
Weaver, D.H. (ed.) (1998). The global journalist: news people around 
the world. New Jersey: Hampton Press
Weaver, D. H., & Wilhoit, G. C. (1996). The American journalists in 
the 1990s: U.S. news people at the end of an era. Mahwah, NJ: 
Lawrence Earlbaum.
White, D. M. (1950). The  "gatekeeper@: a case study in the selection 
of news. Journalism Quarterly, 27, 383-390.


Tables

Table 1: Descriptive Statistics
Percentage
Mean
StD Deviation
N
Age (1=18-24 to 3=34-44 and to 6=65+)
3.69
1.69
987
Sex
Male
46.3
464
Female
53.7
538
Income  (4=30,000-40.000)
4.46
1.62
996
Education (1=below high school, to 5=above college )
4.46
1.624
996
Party ID
Republican
31.6
296
Independent
37.0
347
Democrat
31.4
295
Political View (5=very liberal )
2.77
.968
951
TV News Exposure
(1=rare to 4=regular )
2.97
1.05
997
Believability (1=can't believe to 4= believe most)
CNN
3.03
.915
871
Fox
2.81
.888
781
PBS
2.87
.906
516





Table 2: Zero-order Correlations between Independent Variables and 
Believability .
CNN
Fox
PBS
age
-.152**
.081**
.122**
SEX
.198**
.090**
.146**
Education
-.048
-.132**
.164**
Income
-.051
-.106**
.036
Party ID
.038
.081**
.012
National TV Exposure
.146**
.090**
.083
political views]
.114**
-.132**
.070
*p<.05; **p<.001






Table 3: Believability Regressed on Predictors
Standardized Co-Efficient
CNN
Fox
PBS
Age
-.162*
.024
.143**
Gender
.205**
.079*
.152**
Education
-.054**
-.098*
.183**
Income
-.029
-.097*
-.005
Party ID
.003
-.092*
.013
Cumulative Adjusted R2
.063
045
.070
National TV  News Exposure
.166**
.234**
.079
Cumulative Adjusted R2
.089
.095
.076
Political View
.089**
-.174**
.064
Cumulative Adjusted R2
.095
.121
.077
*p<.05; **p<.001