Perceptions of News Credibility about the War in Iraq: Why War Opponents Perceived the Internet as the Most Credible Medium

Authors

  • Junho H. Choi,

    1. School of Communication Arts
      Kwangwoon University, Seoul
    Search for more papers by this author
    • Junho Choi is an associate professor of Digital Media in the School of Media Communication at Kwangwoon University, Seoul. His research interests include communication networks of online groups, global structure of information flows, and metadata analysis of digital media users and contents.

      Address: School of Communication Arts, Kwangwoon University, Nowon-gu, Seoul, South Korea

  • James H. Watt,

    1. Department of Language, Literature, and Communication
      Rensselaer Polytechnic Institute
    Search for more papers by this author
    • James H. Watt is Professor and Director of the Social and Behavioral Research Laboratory at Rensselaer Polytechnic Institute. He is the author of two books and over 70 research articles, book chapters, technical reports, and papers. His research interests include computer-mediated collaboration and digital games. He is cofounder of a Web survey research firm and has served as a frequent consultant to high technology, market research, and media firms.

      Address: Department of Language, Literature, and Communication, Rensselaer Polytechnic Institute, Troy, NY, 12080 USA

  • Michael Lynch

    1. Department of Language, Literature, and Communication
      Rensselaer Polytechnic Institute
    Search for more papers by this author
    • Michael Lynch is a clinical assistant professor at Rensselaer Polytechnic Institute. His research interests include human-computer interaction, analysis of computer game interfaces, design of AI within computer games in support of social interaction and communication, cognitive processes for modeling computer game AI, and speech act theory.

      Address: Department of Language, Literature, and Communication, Rensselaer Polytechnic Institute, Troy, NY, 12180 USA


Abstract

This study investigated cross-media credibility perception with respect to news coverage about the Iraq War. In an environment of political partisanship, perceptions of media credibility were likely affected by the audience’s political position on the war. Based on hostile media effect theory, a set of hypotheses was proposed to investigate whether the minority opinion group, war opponents, evaluated the Internet as a more credible medium than did neutrals or supporters. An online survey was conducted to which 481 people responded (71% war supporters, 19% opponents, 10% neutrals). Results showed that opponents of the war perceived the Internet as less aligned with a pro-government position and as more credible than did neutrals or supporters. The opponent group also showed a strong negative correlation between perceived pro-government alignment and perceptions of Internet credibility. For the minority partisan group, the diversity of information and views on the war was the main reason for the perception of high credibility of the Internet as a news channel.

News about the war and the dominant medium

During the war in Iraq in 2003, there were more sources of war information available to audiences in the United States than during any previous war. These illustrated a complex array of competing and complementary characteristics of production, distribution, and consumption of news through diverse media. For the first time, the Internet emerged as an important channel of war news for a significant portion of the news audience. Along with conventional mass media, the Web added its interactivity and diversity to the menu of available news channel options that people could access for war information. Given this variety of information sources, how credible did news consumers perceive the Internet to be as a medium for getting news about the war?

Historically, the 2003 war in Iraq can be seen to have occurred within a newly changed media environment. For instance, the choice of dominant medium for obtaining information—and the consistency of its coverage and its alignment with official government positions—has shifted from the Vietnam War to the Gulf War to the Iraq War.

Network television was the dominant medium during the Vietnam War. Television reports used edited film, but from uncontrolled reporting sources. They showed very graphically what was happening on the ground. The discrepancy between the positive government position at the time and the events being shown probably overrode the fact that the coverage was limited in amount and was clearly edited before broadcast. In part for this reason, media coverage was effective in eventually convincing the public that the war was not going well, despite official government pronouncements.

During the Gulf War in 1991, the media were tightly controlled and showed few actual battle operations in real time. Those were, like the well-known video clip of a smart bomb precisely zeroing in on its target, selected because they placed combat operations in the most favorable light in support of the government’s position. During that war, a unique medium for war news emerged in the form of the cable news service, CNN, which provided 24 hours of uninterrupted coverage with global reach (Pan, Ostman, Moy, & Reynolds, 1994). However, because there was little live coverage of actual war operations, CNN’s live coverage gave a sense of distance between the viewers and the actual reality of events on the ground. Although the Gulf War was given saturation coverage, it was based on sources whose exposure to actual operations was clearly being restricted by the military (Fox, 1995).

Much of the coverage presented by stand-up commentators in front of command posts. Coverage gave direct viewing of a few events but many of the events appeared to be happening behind closed doors. There was no discrepancy between what was being shown and the official government position, but since the coverage appeared to be controlled, this in itself probably did not serve to enhance the credibility of the medium of television in covering the war.

In the 2003 Iraq War, saturation coverage appeared to be unedited and happening in real time. As with the Gulf War, the positive government position and the coverage were consistent, but the use of live pictures rather than talking heads gave an appearance of unedited information. In actuality, embedded journalists were very closely controlled in their access to events (Katovsky & Carlson, 2003; Schechter, 2003). Unlike in Vietnam, they had no freedom to go where the action was and their positioning was in fact a form of hidden editing or gatekeeping. It seems plausible that many televised operations and events were staged, as seen in ‘Saving Private Jessica Lynch’ and the destruction of a statue of Saddam Hussein (Kellner, 2004; Rutherford, 2004). The 2003 Iraq War coverage combined the Vietnam War coverage sense of “being there” with the continuous coverage alongside the implicit pro-government position of the coverage of the Gulf War of 1991.

On the other hand, the media environment of 2003 was substantially more complex than during the 1991 Gulf War. Pan-Arab satellite news channels such as Al Jazeera and international news channels like BBC were readily accessible to American news consumers via websites (Hamdy & Mobarak, 2004). These news sources reflected views quite unlike those of the domestic U.S. media, and American online users could access alternative voices from those foreign news sources. In addition, blogs, pro- and anti-war groups’ websites, and emails were actively used to distribute information and opinions. However, it is difficult to declare which medium was the dominant channel for war news. Most of the traditional news organizations extended their services to the Internet as a means of securing additional revenue streams. Non-traditional news sources, namely portal sites and non-branded online news sites, also competed with these traditional news providers.

It seems evident that the Internet had emerged as a major channel for news during the 2003 war. While one-quarter of online users normally get news from the Internet on a typical day, a recent PEW survey found that more than three-quarters of online Americans made use of the Internet in connection with the war in Iraq (Rainie, Fox, & Fallows, 2003). According to that survey, the most popular sources of news were the branded news websites maintained by newspapers and television networks. People reported that the main reasons for using the Internet to get news were for source variety, rapid updates, and different news points of view. However, the study did not investigate how people evaluated the quality of news from different media or how credible they perceived the Internet to be as a channel for news during wartime. To understand better the role of the Internet in times of international conflicts, we ask first how people perceived this new news medium.

Perceptions of online news credibility

People tend to seek more information and up-to-date news in times of emergency and national or international conflicts. The use of news content from various media during periods of war provides an ideal opportunity to study and compare the perception of credibility that news consumers have towards each medium.1

With a main focus on the Internet, this study investigated people’s perceptions of the credibility of media in providing information about the war in Iraq in 2003. Most prior studies of media credibility have asked people to rate the credibility of each medium without specifying the type of news content itself. With only a single news topic as the focus of the present research, this study offers a solid framework for a cross-media comparison of people’s perceptions of news credibility within a context of highly dramatic, (inter)national events.

Has the Internet become a credible channel that supplements or substitutes for traditional news media? The credibility of the Internet in general as a new medium for getting information has been a popular research theme in recent years. Inaccuracy and bias due to the lack of editorial and gatekeeping rules have been suggested as primary reasons why information on the Web is vulnerable to being perceived as non-credible (Abdulla, Garrison, Salwen, Driscoll, & Casey, 2004; Bucy, 2003; Flanagin & Metzger, 2000, 2001; Sundar, 1998, 1999). Certainly, there is some merit to this accusation for general information available on the Internet, but not for most branded news organizations’ websites. It is quite interesting to note that the Internet was judged a more believable medium by a considerable margin when users were asked to compare traditional news sources with their online counterparts, such as CNN with CNN.com (Rainie, Fox, & Fallows, 2003).

The prominence of online news credibility may be explained by a brand name or halo effect, arising from its relationship with existing print and television news organizations. If people evaluate The New York Times as highly credible, they tend to rate NYTimes.com as credible as well. They will also assign some credibility to linked blogs or websites. However, the brand or halo effect does not adequately explain the higher credibility enjoyed by online news outlets, given the similarity of their content to that of traditional media.

Despite the fact that online news sites provide readers with distinctive functions such as flexibility in story selection and interactivity, Web-based news formats with hyperlinks and multimedia have been only cited in terms of having different issue salience than printed newspapers (Althaus & Tewksbury, 2000, 2002). We need to discuss further how the concept of media credibility is defined and how credibility scores are measured for purposes of making cross media comparisons. Also, it is necessary to probe whether the concept and measurement of credibility for traditional news media are still valid and are useful for the Internet. In addition, we need to question whether media credibility is solely an innate characteristic of the channel and source of news, or is in some fashion dependent on audiences’ cognitive processes. The issues of concept, measurement, function, and cognitive effect are discussed in the following section.

Literature review

News credibility and cross-Media comparison

Concept of media credibility

As a factor underlying a reader’s or viewer’s evaluation of news, the concept of credibility has been defined primarily in three ways: (1) message (or story), (2) source (or organization), and (3) media (or channel) credibility (Sundar, 1999). Whereas story credibility focuses on perceived characteristics of news content and source credibility on message senders, media credibility has been defined as the perception of a news channel’s believability (Bucy, 2003; Fico, Richardson, & Edwards, 2004; Kiousis, 2001). Obviously, these three different credibility concepts overlap in part, and it has been argued that all message, source, and channel characteristics constitute the overall perceived credibility of news (Kiousis, 2001).

Cross comparison of media credibility has been a recurring research theme for mass communication scholars and journalists, particularly since Roper polls in the early 1960s showed that audiences evaluated television as a more credible channel for news than newspapers (Gaziano & McGrath, 1986). The credibility of television, newspapers, radio, and magazines has been measured and compared: Television has frequently been named as the most credible medium, but overall credibility ratings of all media have declined over time (Finberg & Stone, 2002).

Measurement of media credibility

The measurement of media credibility has been inconsistent, and different operationalizations of this construct have led to discrepancies in media credibility ratings across studies (Flanagin & Metzger, 2000; Johnson & Kaye, 2002). In the same vein, the measurements of Internet credibility in comparison with other media have certainly also been problematic. There has been a lack of consistency and conclusiveness in credibility studies that compare the Internet to traditional news channels (Althaus & Tewksbury, 2000, 2002; Bucy, 2003; Sundar, 1998, 1999).

When measured as a single dimension, media credibility typically makes use of believability as the surrogate measure (Bucy, 2003). However, media credibility has traditionally been considered a multidimensional construct, although the composition of credibility dimensions has been inconsistent across studies (Gaziano & McGrath, 1986; Kiousis, 2001; Meyer, 1988; Mulder, 1980, 1981). Along with believability, the most common components of media credibility emerging from past studies are accuracy, fairness, lack of bias, completeness, depth, and trustworthiness (Flanagin & Metzger, 2001; Johnson & Kaye, 1998, 2002). However, these may be aspects of credibility more applicable to journalistic practice than to the manner in which news consumers select and judge news channels (Finberg & Stone, 2002).

Multidimensional indices with the same components of credibility evaluation were applied unilaterally to different media, and the index scores of each medium were compared. However, this multidimensional approach to the measurement of media credibility has been criticized on the basis that people may have different sets of credibility perceptions for different news media (Newhagen & Nass, 1989; Rimmer & Weaver, 1987). As Sundar (1999) points out, most studies on the perception of news credibility have not administered open-ended questionnaires to respondents, and little attempt has been made to elicit different elements of credibility perceptions from different media. When people consume different media for the same news topic, they may have different reasons for their selective use of media, which may then lead to different perceptions of relative credibility. For example, readers may expect depth of coverage from magazine news but expect timeliness from news on the Internet. Viewers of television news may regard live pictures as a critical component for credibility while readers of newspapers may regard the expression of diverse opinions on op-ed pages as a critical component for credibility.

Thus, the process of making cross-media comparisons using multidimensional constructs may be flawed. Credibility may not be solely an objective property of the source or attribute of news content, but may also be thought of as the audience’s perception of information from diverse sources of their choice as well as their cognitive processes for assessing news content (Gunther, 1992). From the audience’s point of view, the assessment of news content may be affected by the audience’s different evaluation criteria for different media.

Rather than using a pre-defined multidimensional index, the measurement of audience perceptions of media credibility may be made more precise if respondents are asked to rate the relative credibility of media of their choice. By using a single event for relative assessment of media, measurement of credibility may more properly reflect the audience’s perception of media characteristics when they assess news content, because news consumers have different expectations and context for use of each different medium.

Function of media credibility

What is the function of perceived media credibility on an audiences’ news consumption process? Several demographic characteristics of media users, in particular, gender, age, income, and education, have been identified as important predictors of media credibility.2 However, these findings on demographics are mostly atheoretical. In their exploration of agenda setting effects, Wanta and Hu (1994) proposed a model in which high credibility perception leads to reliance on a certain medium for news, and strong reliance, in turn, increases exposure to the medium. Reliance is an attitudinal value towards media on which audiences depend for gaining information (Johnson & Kaye, 2000, 2002). If news audiences perceive the Internet to be highly credible as a news medium, they become highly dependent on that medium. However, credibility perception has shown a weak association with newspapers and television use (Rimmer & Weaver, 1987; Wanta & Hu, 1994). These findings from past studies, while suggestive, have not have been sufficient to establish the relationship between reliance and media credibility firmly for all media.

With regard to the Internet, the causal relationship between credibility perception and reliance may be reversed. Johnson and Kaye (2000, 2002), for instance, found that reliance was strongly associated with credibility perception of online media sources during the presidential campaign, but negatively predicted online newspaper credibility. The Internet is a versatile medium in which users can exploit a number of diverse communication activities, ranging from interpersonal correspondence to sampling a variety of information sites and to participating in public discussions. The context of use of the Internet may be quite different from that of other media such as television. For example, people may come to rely on the Internet for access to news when other media are not available, for instance, while at the workplace. They may also depend on the Internet when they want to access a news source for a topic of their choice at a time of their choosing. Also, supporters and opponents of the war might have used the Internet more aggressively in order to form alliances and to share opinions in support of their respective positions.

Hostile media effect: Issue-involved audiences and media bias assessment

Political attitudes towards the war may have strongly influenced media use and credibility perception of news coverage. The war in Iraq in 2003 provoked a split between supporters and non-supporters domestically as well as globally. When there is a strong political partisanship, the perception of media credibility may be affected by the audience’s issue involvement in a specific news event. There is an emerging perspective on media credibility studies that news credibility may be a factor in audience involvement in the issue, rather than internal or external characteristics of news content itself. According to this perspective, news credibility is not solely an objective feature of news story or source, but a subjective perception by audiences.

The hostile media effect refers to the process by which some news audiences, when they are highly involved in an issue, tend to perceive balanced and neutral news coverage as biased against their own point of view and judge the news content as less credible (Arpan & Raney, 2003; Gunther, 1992; Gunther & Chia, 2001; Gunther & Christen, 2002; Gunther, Christen, Liebhart & Chia, 2001; Vallone, Ross, & Lepper, 1985). That is, when two opposing partisan groups are exposed to presumably neutral coverage on a divisive issue, both groups will evaluate the identical news content as unfavorable towards their position and, thus, as unbalanced and biased. Researchers have examined this effect between diverse partisan groups when both were exposed to the news coverage on the involved issue: pro-Arab and pro-Israeli groups (Vallone, Ross, & Lepper, 1985), managers and union members on strike (Christen, Kannaovakun, & Gunther, 2002), and opponents and supporters of genetically processed foods (Gunther & Schmitt, 2004).

The hostile media effect assumes two theoretical conditions: a highly issue-involved audience and neutral news content. A collection of people having strong involvement with a social or political issue is sometimes referred to as a partisan group. When news consumers have strong opinions towards an issue, they are more likely to express skepticism of the news coverage, and the magnitude of their attitude shows positive correlation with the degree of bias perception (Arpan & Raney, 2003; Gunther, Christen, Liebhart, & Chia, 2001; Schmitt, Gunther, & Liebhart, 2004). The partisanship is often identified by membership in a social or political group or organization.

A fundamental ethic in journalism is the goal of writing a neutral or balanced news story in which both sides are equally well covered. However, it is quite common for news editors to receive complaint letters from readers on both sides of an issue to the effect that news coverage is biased in favor of the other side. Individuals’ information processing mechanism may be explained by several different theories. One of these, hostile media perception, seems to be related to third-person effect and mass-reach effects. Another theory, the assimilation bias effect, is a classic and traditional explanation of biased cognitive processing theory that refers to the tendency of people with opposing views to evaluate inconclusive information in favor of their own opinions.

In contradistinction to this supportive perception of assimilation bias theory, hostile media effect theory predicts opposed perception, but the prediction is found to be accurate only if a medium is “broad-reach” (Gunther & Schmitt, 2004). It is the perception of a medium having a broad-reach capacity, with its concomitant potential influence on large numbers of others (especially neutral opinion groups) that generates the hostile perception. As proposed by the third-person effect (Davison, 1983; Perloff, 1999), when partisan groups are exposed to nominally neutral news coverage in mass media, they tend to perceive it as having a larger undesirable influence on others than on themselves and, thus, perceive the news as unfavorable to their side. When they are exposed to news content under non-mediated or narrow-reach conditions, the hostile media effect disappears and assimilation bias comes into play (Gunther & Schmitt, 2004).

The mechanisms through which the effect occurs involve judgment heuristics: Partisans who believe their camp is superior to the opposing side may judge neutral news as hostile because such a balanced coverage gives equal weight to the other side’s inferior claims. When war supporters sensed their dominance in opinion polls, they would tend to judge a seemingly balanced report as biased because they might think it favored the inferior group. Also, partisans who believe a news source is aligned with the other side constantly assume bias in all news coverage from the news source. For instance, war opponents would have assumed bias in news coverage from Fox News if they believed that news source always favored the other side.

These rather strict conditions on the neutral coverage and partisan groups have been softened in later studies. Relative hostile media effect theory argues that the news content does not need to be neutral. Even if news coverage is slanted, both partisan groups will evaluate the news content as relatively biased against their side (Gunther & Chia, 2001; Gunther & Christine, 2002). For instance, even if news coverage is slanted towards one side, the group will still perceive the news as unfavorable to their position, although their perception of bias is relatively weaker than the other side’s. This modification eliminates the requirement that the hostile media effect (HME) be tested only on neutral news coverage. In actuality, there is no guarantee of perfect objectivity in most news coverage; in the experimental examination of HME, researchers have often manipulated news stories in order to meet the neutrality requirement.

In addition, relative HME argues that the contrast bias effect is not limited to extreme news audiences, that is, highly issue-involved partisans who are usually only small subsets of the general population. Although only moderate correlations were found, researchers have also identified the perception of bias in news coverage toward the other side in broader samples of the population (Gunther & Christine, 2002).

Supporters and opponents of the Iraq War were markedly partisan groups who were highly involved in this issue. Even if the coverage in a certain medium they were exposed to in some sense tended towards neutrality, they would nevertheless have judged that medium as biased, thus giving it a lower credibility assessment than would neutral groups. However, in a non-experimental study, it is difficult to determine whether the news coverage they were exposed to was neutral or not. When a slant or bias by media is present, proponents of one side of an issue who are exposed to news unfavorable to their side tend to assess the news coverage or media as more slanted or biased than do proponents of the other side (Arpan & Raney, 2003).

Thus, people’s perception of each medium in terms of the degree of its alignment with the government’s position may be an important factor in credibility perception during the war. The government officially claimed the weapons of mass destruction and a link to al-Queda as major reasons for the war, but provided little solid evidence. If one is in support of the war and the government’s position, then a medium that is perceived as supporting that position may be judged as more credible than other media. If one does not support the war or the government’s position, then a pro-government medium may be seen as less credible than other media. Thus, both effects of audience’s issue involvement and the medium’s alignment with the government’s position on the assessment of credibility ought to be considered. For instance, if opponents of the war had perceived most news sources on television to be pro-government, they would have given lower evaluation of credibility of the medium of television overall.

Rationale and hypotheses

The main objectives of this study were to examine the hostile media effect and to investigate perceptions of credibility of the Internet. War opponents in times of war are a minority partisan group whose political attitudes are not in accord with the government’s position. Thus, if they perceived that news sources from mainstream media could not provide the information and commentaries they sought, they would rely more on alternative channels and judge those channels as more credible.

We used the 2003 war in Iraq as a single news event for the examination of the hostile media effect on news credibility perception. The news about the war provided a unique opportunity for understanding the Internet’s role in reporting on a significant international issue and for understanding the hostile media effect on the minority (anti-war) opinion group. For the study of issue-involved audiences’ perception of news, previous studies adopted diverse news events, but the range of issues tended to be domestic (e.g., a presidential election), local (e.g., local sports news coverage), or non-general (e.g., genetically processed food). The war in Iraq was an international event and many people used (and were able to use) the Internet to access non-domestic news sources such as BBC and Al Jazeera. The war in Iraq attracted large news audiences, and also provoked a sharp split between supporters and non-supporters. Given that strong issue involvement is a key assumption of hostile media effect theory, the news about the war was ideal for examining the effect of a partisan audience on news credibility perception. Furthermore, unlike the composition of opinion groups on such events as presidential elections, there was a sizable disparity between majority and minority opinion groups, which made it interesting to examine how audiences on the side of minority opinion used and evaluated the news on the Internet.

Three hypotheses were proposed to test the effects of issue involvement and media assessment on news credibility perception in the context of the Iraq war.

H1: War opponents will perceive the Internet as less aligned with the government’s position than will war supporters.

H2: War opponents will rate the Internet as a more credible medium than will neutrals and supporters.

H3: There will be a strong negative correlation between perceptions of credibility and perceptions of pro-government alignment of the Internet among the war opponents.

Methodology

Sample and procedure

The online survey was conducted shortly after most military battles had concluded during the Iraq War in April 2003. The sample for this study was obtained from a commercial online survey firm. Respondents were randomly selected from a pre-qualified panel that was demographically representative of all users of the Internet in the U.S. The sampled individuals indicated that they resided in the U.S. and were older than 18 years. The sampled individuals were sent an email recruiting message with a link to the online survey system that presented the questionnaire.3

The survey asked respondents to furnish demographic information, attitude towards the war, frequency of media use for war information, perceptions of credibility of media, and the perception of each medium’s alignment with the U.S. government’s position. Also, respondents were asked to answer an open-ended question about the reasons for their perception of the credibility of each medium.4

A total of 528 responses were received. After discarding 47 cases that reported no use of television or the Internet for war information, a total of 481 cases were used for analysis. The resulting sample was 51% male and 49% female. The average age was 38.9 (s.d. = 10.9). The average number of education years was 14.9 (s.d. = 2.0), and average income was $57,505 (s.d. = 22,758). Seventy-one percent of the respondents supported the war, while 29% did not support it (19% were opposed and 10% were neutral).5 These figures agreed very closely with national survey percentages obtained at the same time from samples of the online population (74% of whom were supporters; Rainie, Fox, & Fallows, 2003) and the general population (69% of whom were supporters; CBS News, 2005).

Measurement

Attitude towards the war

Respondents were asked to indicate whether they opposed, were neutral or undecided towards, or supported the war in Iraq.

Media use

In most previous studies, media use has been measured in three different ways: retrospective reports of use time (i.e., how many hours of use per day), yesterday’s use of the medium (i.e., used it or not yesterday), or frequency of use (i.e., how many days per week). Given the short duration of the coverage of active war news at the time of the survey and the mixture of various media use at several times a day that news consumers would likely avail themselves of for such a dramatic and breaking news event, we felt that a short-term frequency of media use measure would yield better recollection scores than a media use time item and a wider variance for statistical analysis than yes/no data from a “used it or not yesterday” item. Thus, we concluded that a short-term frequency of media use measure would be the most appropriate. Frequency of use was measured by asking “During the past week, how often did you get news about the war from each medium?” The reported frequency was recalculated into average daily frequency of use for each medium.

Medium’s alignment with the government position

Media use and the perceived credibility of a medium might be affected by the perception of its alignment with official government positions. Respondents were asked to rate their agreement with the question, “Overall, the information available on {the Internet, television, newspapers, magazines, radio} is very consistent with the government’s position” using a seven-point Likert scale, with higher numbers indicating more agreement with the statement.

Media credibility

Initially, respondents were asked to rank-order all media on overall credibility for war news, regardless of whether or not they used the medium. We then selected for final analysis those subjects who used both the Internet and television for war news. Credibility was assessed by asking the respondent to “Rank all media on overall credibility for war news.” Because many studies argued that previous measurement schemes of media credibility were problematic, we chose relative evaluation of five media with a single dimensional question over the assessment of each medium with the same pre-defined multidimensional index. For purposes of investigating the relative hostile media effect, a measurement of audience perception of media credibility can be made more precise when respondents are asked to rate relative credibility of all available media and rank them in order. The ranking given to each medium was converted into a credibility score by subtracting the ranking from 6, giving a credibility score that ranged from 5 (most credible) to 1 (least credible). Thus, the credibility score for each medium is the reversed ranking of the medium out of the five media presented.

Results

Alignment of the internet with government position

In order to examine if opponents rated the Internet as less aligned with the government’s position than did supporters or neutrals, we analyzed how each group perceived each medium’s alignment with the government’s position. Table 1 presents the mean scores of perceptions of media alignment with the U.S. government position by war opponents, neutrals, and war supporters.

Table 1.  T-test results on the perception of media alignment with government position by war opponents, neutrals, and supporters
 Opponents (n = 89)Neutrals (n = 50)Supporters (n = 342)
  1. Notes: *Indicates that the mean difference from opponents is significant at the .05 level. **Indicates that the mean difference from opponents is significant at the .01 level.

Internet3.69 (1.63)4.46 (1.39) **4.68 (1.47) **
Magazines4.87 (1.23)4.66 (1.12)4.13 (1.38) **
Radio5.01 (1.34)4.80 (1.18)4.62 (1.34) *
Newspapers5.22 (1.25)4.98 (1.12)4.13 (1.38) **
Television5.88 (1.17)5.22 (1.28) **5.07 (1.59) **

Opponents rated television as the medium most aligned with the government’s position (m = 5.88, s.d. = 1.17), followed by newspapers (m = 5.22, s.d. = 1.25), radio (m = 5.01, s.d. = 1.34), magazines (m = 4.87, s.d. = 1.23), and the Internet (m = 3.69, s.d. = 1.63). The Internet was perceived as the least pro-government channel for war information. Supporters rated television (m = 5.07, s.d. = 1.59) as most aligned with the government position and perceived newspapers (m = 4.13, s.d. = 1.38) and magazines (m = 4.13, s.d. = 1.38) as the least pro-government media. The Internet (m = 4.68, s.d. = 1.47) came in second in terms of its perceived pro-government alignment, among supporters. Neutrals evaluated television (m = 5.22, s.d. = 1.28) as the most pro-government channel for war information and the Internet (m = 4.46, s.d. = 1.39) as the least.

All three groups (opponents, supporters, and neutrals) perceived television as the most pro-government medium. However, supporters and neutrals rated the degree of alignment as less strong than opponents did. The mean difference of television’s alignment with government position between opponents (m = 5.88, s.d. = 1.17) and supporters (m = 5.07, s.d. = 1.59) was significant (t = 4.47, d.f. = 183.22, p < .01). The t-test difference between opponents and neutrals (m=5.22, s.d.=1.59) was also significant (t= 3.07, d.f. =137, p < .01).

Opponents (m = 3.69, s.d. = 1.63) rated the Internet as less aligned with the government position than did neutrals (m = 4.46, s.d. = 1.39) and supporters (m = 4.68, s.d. = 1.47). Independent t-tests were run to test if the mean differences between opponents and the other groups were random. The result showed that there was a significant difference between opponents and supporters in perception of the Internet in terms of alignment with government position (t =−5.53, d.f. = 429, p < .01). The t-test result was also significant between opponents and neutrals (t=−2.83, d.f. = 137, p < .01).

Perception of credibility of the internet

The second hypothesis asked if there was a significant difference in the credibility perception of the Internet between opponents and other groups. Credibility perception scores on five media by three groups are presented in Table 2.

Table 2.  T-test results on the perception of media credibility by war opponents, neutrals, and supporters
 Opponents (n = 89)Neutrals(n = 50)Supporters (n = 342)
  1. Note: *Indicates that the mean difference from opponents is significant at the .05 level. **Indicates that the mean difference from opponents is significant at the .01 level.

Internet3.58 (1.36)3.36 (1.36)3.30 (1.30)
Newspapers3.24 (1.22)3.28 (1.06)3.15 (1.11)
Television3.13 (1.51)4.08 (1.20) **4.13 (1.21) **
Radio2.59 (1.26)2.58 (1.05)2.66 (1.16)
Magazines2.35 (1.33)1.84 (1.26) *1.74 (1.06) **

Opponents rated the Internet (m = 3.58, s.d. = 1.36) as the most credible medium, followed by newspapers (m = 3.24, s.d. = 1.22), television (m = 3.13, s.d. = 1.51), radio (m = 2.59, s.d. = 1.26), and magazines (m = 2.35, s.d. = 1.33). Neutrals chose television (m = 4.08, s.d. = 1.20) as the most credible channel for war news; the Internet (m = 3.36, s.d. = 1.36) was second. Supporters also perceived television (4.13, s.d. = 1.21) as the most credible medium, followed by the Internet (m = 3.30, s.d. = 1.30).

A series of independent sample t-tests were run in order to check whether opponents rated the Internet as a more credible medium than other groups did. The Internet was rated as more credible by opponents (m = 3.58, s.d. = 1.36) than by neutrals (m = 3.36, s.d. = 1.36) or by supporters (m = 3.30, s.d. = 1.30). However, the independent t-tests showed that the mean differences were not significant.

The credibility scores for television of supporters (m = 4.13, s.d. = 1.21) and neutrals (m = 4.08, s.d. = 1.20) were higher than that of opponents (m = 3.13, s.d. = 1.51). The differences between supporters and opponents (t = 6.51, d.f. = 428, p < .01), and between neutrals and opponents (t = 3.77, d.f. = 121.23, p < .01), were both significant.

Correlation: Media credibility and government alignment

The third hypothesis proposed that there would exist a strong negative correlation between perceptions of credibility and perceptions of pro-government alignment of the Internet among war opponents. Correlation analyses were run in order to find evidence of this association; the results are presented in Table 3.

Table 3.  Correlations of perception of pro-government alignment and perception of media credibility by opponents, neutrals, and supporters
 Opponents (n = 89)Neutrals (n = 50)Supporters (n = 342)
  1. Note:*p <.05, ** p<.01

Internet−.49 **.00.14 **
Television−.25 **−.18.41 **
Newspapers−.17.20.22 **
Radio−.20−.04.22 **
Magazines−.01.28.06

Among opponents, a strong negative correlation was found between perception of pro-government alignment and perception of credibility of the Internet (r =−.49, p < .01). This suggests that opponents evaluated the Internet as a more credible channel for war news and information because they believed that they could find more news and information on the Internet not aligned with the pro-government position.

By contrast, there was a positive correlation between perception of pro-government alignment and perception of credibility of the Internet among supporters (r = .14, p < .01), as well as a strong positive correlation between perception of pro-government alignment and perception of credibility of television (r=.41, p < .01). This suggests that supporters evaluated television as a credible channel because they could find news and information there that are compatible with their attitudes towards the war and the government’s position. The neutral group did not show any significant correlation between media credibility and pro-government stance for any media.

Discussion

The main objective of this study was to investigate the extent to which American news audiences perceived the Internet as a credible news medium during the Iraq War. As one of the fundamental criteria by which people judge the quality of the news they receive, credibility is not merely a characteristic of the message itself, but a perception by the audience. When dealing with news content concerning a salient issue, issue-involved news consumers tend to judge media credibility differently.

This study questioned whether audiences held different perceptions of media as a function of their attitudes towards the war. The Iraq War provided an ideal situation in which to probe the relationship between an issue-involved audience and their perceptions of credibility of different news media. Three groups were identified by their attitude towards the war in Iraq: opponents, neutrals, and supporters.

The first hypothesis proposed that opponents would rate the Internet as less aligned with the government’s position than would supporters or neutrals. This hypothesis was accepted: There were significant differences between opponents and other groups in the perception of the Internet as an alternative medium for war news. Opponents of the war sought information and news from the Internet that provided diverse points of view in accordance with their stance.

The second hypothesis proposed that opponents would rate the Internet as a more credible medium than non-opponent groups. Partial support for this hypothesis was found: Opponents of the war perceived the Internet as a more credible medium for war news and information than did neutrals or supporters. However, the mean differences failed to reach a significance level in the sample.

The third hypothesis proposed to investigate the association between pro-government alignment perception and credibility perception of the Internet by opponents. Strong support for this hypothesis was found: Opponents showed a strong negative correlation between the perception of credibility and perceived pro-government alignment of news sources on the Internet. The existence of diverse sources and anti-war views from the Internet was the main reason for opponents evaluating the Internet as a more credible medium.

Thus the Internet was viewed as an alternative channel by those who could not otherwise get news or information from traditional media that accorded with their political attitudes towards the war in Iraq. Opponents of the war demonstrated a distinctive perception of the Internet, judging the Internet as less aligned with the government’s position as well as a highly credible channel for war news. As described in note 4, open-ended responses to the question of “main reason for choosing a specific medium” clearly revealed that news consumers have different criteria for credibility assessment of each medium. The dominant criterion for the Internet was diversity in the sources of information, while for television it was the availability of live visual images. Results from this question indicated that, from the audience’s point of view, applying uniform a priori dimensions to assess each medium’s credibility lacks validity.

Conclusions

These findings provide insights into fundamental issues regarding the definition, cognitive processing, and measurement of the credibility of news media construct. Is credibility an objective quality of news? Are there universal criteria by which to judge the level of news credibility? Hostile media effect theory provides a solid theoretical framework to shed light on the mechanism of differential news perception by issue-involved audiences. As suggested in this study, whenever a salient issue creates a highly partisan split between supporters and opponents, news credibility is both a subjective perception by audiences and a function of their cognitive processing mechanisms, rather than simply an innate quality of news stories or sources themselves. Unlike previous studies on hostile media effect theory, this study demonstrated evidence that the HME continues to hold even when the sizes of partisan groups differ and when the issue at stake engages general news audiences.

The results of this study of media credibility may also have consequences for journalistic practice within the news industry during a period when the media environment is changing rapidly due to the continued adoption of the World Wide Web (and other Internet resources) as part of people’s news gathering strategies. What does balance in news reporting mean when people can access a virtually limitless number of news sources? When an issue splits the public into partisan groups, how can news organizations be judged on the criterion of adhering to a balanced or objective position? The findings of this study suggest that news producers cannot strike a balanced stance simply by steering a median or centrist course. For many salient news topics, the emerging relative sizes of opposing and supporting audiences may differ, and this degree of difference may accelerate the so-called ‘spiral of silence’ effect on public opinion for whichever side has the smaller audience. Researchers may focus on the neutral or undecided group and ask how they evaluate the news or news organizations, because issue involvement itself has an impact on how partisan viewers in either camp evaluate balance and credibility.

This study was conducted under natural field conditions. Its main strength lies in the fact that we tested the effect of issue involvement of news audience groups in a non-experimental setting. Thus, we felt justified in drawing inferences about news credibility and the role of the Internet in real-world conditions.

The weakness of this study stems from the same condition. Because we could not control messages, exposure, or any other external conditions, we have only limited evidence of the effect of issue involvement on the perception of media credibility. The process of news credibility perception could not be structurally investigated. Most findings were based only on group comparisons and bivariate correlations.

Some limitations of this study could be addressed in future studies. We did not explore how people used interactive tools such as email, blogs, and discussion boards in order to share information and views. Future researchers could design studies to capture how those various interactive uses of the Internet might also affect the perception of news credibility.

In addition, researchers may wish to investigate how partisan and neutral groups evaluate news organizations during times of war. From various studies of the hostile media effect and issue-involved audiences, we can presume differential perceptions of source credibility between neutral groups and issue-involved groups. Neutrals would tend to rate media organizations as more credible than either supporters or opponents. Pro-government and conservative sources would be evaluated as more credible by supporters than by neutrals or opponents. Similarly, groups opposed to predominant pro-government positions about a topic would regard anti-government or alternative sources as being more credible.

Notes

  • 1

    Studies of war news and audiences’ perception of media credibility are rare in the literature. Some have reported increased use of media for getting news during times of war, but findings on credibility have not been consistent. Pan, Oatman, Moy, and Reynolds (1994) examined the relationship between level of news exposure and knowledge acquisition for U.S. citizens during the first Gulf War. They discovered increased news media exposure during the war and that newspapers were the most effective medium for knowledge dissemination. Lo (1994), in a study of Taiwanese students, found that newspaper use was more strongly correlated with knowledge acquisition than was television use under high involvement conditions. Saudi-Arabian citizens reported that they relied more heavily on broadcast media than on print media, and urbanites judged international news media more credible than did rural dwellers (Al-Makaty, Boyd, & Tubergen, 1994). A study of UAE citizens’ perception of news media credibility for Iraq War coverage showed that media use increased and television was seen as the most credible medium (Ayish, 2004).

  • 2

    Several demographic characteristics of media users have also been identified as important predictors of media credibility, in particular, gender, age, income, and education. In general, female, younger, lower income, less educated news readers tend to evaluate the media as credible while male, older, higher income, more educated audiences are more likely to be critical of the media (Bucy, 2003). The dominant profile of Internet users used to be male, young, and highly educated. However, as the online population increasingly comes to resemble the population at large, demographics may not be strongly correlated with perceptions of online news credibility (Johnson & Kaye, 1998, 2002). The relationship between credibility and demographic variables turns out to be rather complicated for Internet users. A recent study shows that younger news audiences regard both online and TV news as more credible than do older audiences (Bucy, 2003).

  • 3

    A bulk email recruitment letter was sent out to the panel maintained by the commercial online survey firm. Respondents who replied to it clicked on a link within the body of the email, which took them to the questionnaire site. This process continued until the desired quota was reached. People who responded to the email after the quota was reached were then politely turned down. This was necessary because the survey firm pays each respondent with redeemable points, and we sought to limit this number (and therefore our costs) to 500 respondents. The overall response rate to the initial email recruitment was not available because the number of responses after the quota was reached was not furnished to us by the sample supplier.

  • 4

    To investigate the argument that people may have different credibility evaluation criteria for different media, open-ended responses to the question “Why did you choose this medium as the most credible” were analyzed. All responses were scanned for major themes, and 16 categories emerged (15 specific themes plus an “other” category). The first and the second author independently coded the open-ended responses for the presence of these categories. Coder agreement on presence of the themes in the categories was 87.9%, and only 4% of the responses fell in the catchall “other” category.There were large differences among media on the most frequent reason given for choosing a particular medium as most credible. For those who named the Internet as the most credible source, “access to many sources of information” (43.7%) was given as the most frequent reason for high credibility by a factor of almost two over the next most frequent reason, “timeliness of the information” (25.2%), followed by “world viewpoint” (16.0%), “unedited content” (11.8%), “easy access” (7.6%), and “politically independent” (5.9%). The dominant reason for choosing television as most credible was the “feeling of seeing events live and directly” (60.4%). Interestingly, only 17% mentioned the “timeliness of the information available” as a second reason for television’s credibility. “Having time for analysis of information” was the most frequent reason given for believing that the print media were most credible. This reason was given by 30.9% of those naming newspapers and 50.0% of those naming magazines. Compared to other media, a wider range of reasons for believing newspapers most credible was given by respondents. These included the presence of editors to check facts (16.2%), depth of information (13.2%), and unsupported trust in the medium (22.1%). The most frequent reason given for selecting radio was “unsupported trust.” Over half of the respondents (59.5%) who selected radio simply said it was credible, without elaborating on the characteristics of the medium that made it so.

  • 5

    In this survey, females (69.3%) were somewhat less likely to support the war than males were (73.7%), but the difference was not statistically significant. When the attitude toward the war was considered as a contiguous variable and recoded as an interval scale (opposed = 0, neutral = 1, supported = 2), age and education showed modest correlations with support for the war. Older and more educated citizens tended to support the war less than younger and less educated ones (r = .06, p < .05; r = −.17, p < .01).Frequency of use and demographic variables were correlated. Demographic variables showed a different pattern of relationships with Internet use and credibility perception than they did with television. Gender and income were not associated with Internet credibility. People with more years of education perceived the Internet as having less credibility (r =−.17, p < .001), while older users rated the Internet as slightly more credible than younger users (r = .09, p < .05). Age was positively correlated to the perception of television credibility (r = .12, p < .01) while education was negatively correlated (r = −.22, p < .001). Gender was mildly associated with the credibility rating of television. Females had higher credibility perceptions than did males (Point-biserial r = .12, p < .01).

About the Authors

  1. Junho Choi is an associate professor of Digital Media in the School of Media Communication at Kwangwoon University, Seoul. His research interests include communication networks of online groups, global structure of information flows, and metadata analysis of digital media users and contents.Address: School of Communication Arts, Kwangwoon University, Nowon-gu, Seoul, South Korea

  2. James H. Watt is Professor and Director of the Social and Behavioral Research Laboratory at Rensselaer Polytechnic Institute. He is the author of two books and over 70 research articles, book chapters, technical reports, and papers. His research interests include computer-mediated collaboration and digital games. He is cofounder of a Web survey research firm and has served as a frequent consultant to high technology, market research, and media firms.Address: Department of Language, Literature, and Communication, Rensselaer Polytechnic Institute, Troy, NY, 12080 USA

  3. Michael Lynch is a clinical assistant professor at Rensselaer Polytechnic Institute. His research interests include human-computer interaction, analysis of computer game interfaces, design of AI within computer games in support of social interaction and communication, cognitive processes for modeling computer game AI, and speech act theory.Address: Department of Language, Literature, and Communication, Rensselaer Polytechnic Institute, Troy, NY, 12180 USA

Ancillary