Two articles summary

7

Media Coverage of Muslims, Perceived Threats, Ethnocentrism, and Intercultural Contact: Applying Cultivation Theory, Integrated Threat Theory, and the Contact Hypothesis

Zeeba Khan was previously a professor in the Department of Human Communication Studies, California State University, Fullerton. She can be reached at 53 Cairnhill Road, Cairnhill Plaza #10-02, Singapore 229664, or at [email protected]. Jon Bruschke, Ph.D., is a professor in the Department of Human Communication Studies, California State University, Fullerton, Fullerton, CA 92831, USA.

This study examined reported media exposure, perceived threats from Muslims, ethnocentrism, and intercultural contact using cultivation theory, integrated threat theory, and the contact hypothesis. Ethnocentrism was the largest single predictor of perceived threat, accounting for 42% of the variance in perceived threat. To the extent that minority cultural members are perceived as a threat, they are viewed by majority cultural members in negative ways. Perceived threat was a significant antecedent to contact, with knowledge of Muslim culture ac- counting for 16% of the variance in contact. This investigation concludes that the solution to cultural conflict is an increase in intercultural contact. Keywords: Perceived threat, intercultural contact, ethnocentrism, media, Muslims

THE NoRTHwEST JoURNAl oF CoMMUNiCATioN, Vol. 44, No. 1

This article examines data collected in 2007, a unique historical moment when the immediate trauma from the September 11, 2001, attacks had subsided but before subsequent events (U.S. involve- ment in the conflict in Iraq, the engagement in Afghanistan, the assassination of Osama Bin Laden, the so-called Arab Spring of 2011) would refine understandings of terror and its relation to Islam.

zeeba khan and jon bruschke

California State University, Fullerton

The Northwest Journal of Communication_Interior.indb 7 5/3/16 4:42 PM

8 KHAN AND BRUSCHKE

In short, it explores how attitudes about Muslims were formed at a time when world events were still unfolding and many questions about the meaning of September 11, 2001, remained open.

U.S. reaction at the time came in the form of the Bush admin- istration’s package of rhetoric and policy it deemed the War on Terror. Although mainstream media discussions have generally focused on the effectiveness the approach has had on reducing the terror threat, an issue significant in its own right is the effect of that discourse on American Muslims. Major world events can trigger feelings about and influence perceptions of other ethnic, cultural, or racial groups (Sheridan & Gillett, 2005), and there is a high correlation between international or domestic crises in which U.S. citizens are perceived to be the victims of foreigners (and foreign aggression) and an increase in hostility toward non- Caucasian or non-Christian people in the United States (Akram, 2002). Other research has by now begun to explore the influence of the post–September 11 discourse on attitudes about Muslims (Cho et al., 2003; Elasmar, 2008; Nisbet, Ostman, & Shanahan, 2009; Pintak, 2006; Scheufele, Nisbet, & Ostman, 2005; Woods, 2007).

Three theories are especially relevant when seeking to explain the relationship between world events and domestic perceptions of minority groups. First, cultivation theory would predict that ex- tended exposure to media images that depict Muslims in a negative light will result in negative attitudes about Muslims. Second, the specific negative image of Muslims is generally that of a threatening other, making integrated threat theory (ITT) a useful framework for explanation. Third, the contact hypothesis can explain what happens when individuals holding opinions based largely on media images come into direct personal contact with Muslims.

In particular, this project has three goals. First, it seeks simply to explore attitudes about Muslims at a unique historical moment. Most broadly, we are interested in how cultivation theory, ITT, and the contact hypothesis are able to explain attitudes about Muslims

The Northwest Journal of Communication_Interior.indb 8 5/3/16 4:42 PM

9Media Coverage of Muslims

in the wake of the September 11 disaster. Second, this project seeks to extend the theories with traditional theory-building mechanisms, in particular, the addition of variables that are relevant to the processes of the various theories but that have not, to date, been included. Third, this project seeks to integrate the different theories. All theories explain an important component of the overall attitude formation process, but bringing them to bear on this specific context (attitudes about Muslims in the post–September 11 world) requires integrating the theories. Most broadly, we begin by exploring the notion that media exposure produces a perceived threat, which in turn predicts the willingness of non-Muslim Americans to engage in direct contact with Muslims.

LITerATUre reVIew ANd HyPoTHeses

Cultivation Theory and Media Exposure Many studies have examined the effects of exposure to media. Gerbner, Gross, Morgan, and Signorielli (1994) explained that cul- tivation theory is based on findings that television shapes viewers’ perceptions of reality. A classic finding concerns the fear of crime: Individuals exposed to reports of crime perceived themselves to be at a high risk of being victimized by criminal activity (Bryant, Carveth, & Brown, 1981). Although the media is intended to inform, it might inadvertently be creating an environment of fear (Kellner, 2002; Romer, Jamieson, & Aday, 2003). Fear, in turn, can result in a stronger identification with one’s in-group (i.e., increased ethno- centrism) and increased hostility and aggression toward members of the out-group (Argyrides & Downey, 2004).

The fear generated by terrorism has been linked to Muslims. Most generally, Western knowledge of Islam continues to be driven by the distorted writings of early Orientalists (“Western Media,” 2005). Edward Said (1979) explained that canonical Orientalism is framed by political forces that brought the Orient into Western

The Northwest Journal of Communication_Interior.indb 9 5/3/16 4:42 PM

10 KHAN AND BRUSCHKE

consciousness and the Western empire. The Orient is seen as static, homogenous, eccentric, and backward, with a tendency toward despotism and away from progress. It is the “Other”—a mirror image to the West, representing what is inferior, conquerable, and alien (Said, 1979). Print media, radio, and television have perpetu- ated anti-Muslim sentiment (Shaheen, 2001). The predominant stereotypical Muslim men are terrorists and oil sheikhs, whereas the predominant stereotypical Muslim women are scantily clad, exotic belly dancers in harems or weak, mute, faceless creatures enveloped in veils (Akram, 2002). “The terms Islamic or Muslim are linked to extremism, militant, jihads, as if they belonged together inextricably and naturally (Muslim extremist, Islamic terror, Islamic war, Muslim time bomb)” (“Media,” 2005, para. 4; Woods, 2007). Not surprisingly, there has been a notable increase in violence against Muslims and anti-Muslim attitudes since September 11, 2001 (Prengaman, 2006; “Western Media,” 2005).

Researchers have argued that an exaggerated level of public concern about terrorism pushed government officials to support the restriction of civil liberties and that the media played a significant role in creating and sustaining these concerns (Cho et al., 2003; Nisbet et al., 2009; Scheufele et al., 2005; Sunstein, 2003, 2004). These conclusions dovetail with other research that has shown that Americans are far more supportive of restrictions that affect only disliked out-group members rather than the public at large (Mar- cus, Sullivan, Theiss-Morse, & Wood, 1995). The USA PATRIOT Act, which increased law enforcement powers, reduced the flow of information and limited other civil liberties (Woods, 2007). These measures were directed mostly at out-group members, mainly men of Muslim or Arab background.

Cultivation theory is a useful rubric to explain these forces. The general public’s basic knowledge and familiarity with Islam have not increased considerably since 2001 (Nisbet et al., 2009). As such, the general public is heavily dependent on media content to make

The Northwest Journal of Communication_Interior.indb 10 5/3/16 4:42 PM

11Media Coverage of Muslims

evaluations regarding Muslims, and heavy viewers of television news are more likely to agree to greater restrictions on civil liber- ties because of heightened fears of terrorist attacks from Muslims (Scheufele et al., 2005). There can be little doubt that media coverage of Muslims is at least as negative and threat invoking as coverage of crime. To replicate Scheufele et al.’s test of cultivation theory, a first hypothesis is offered:

h1: Greater reported exposure to media coverage will be associated with greater degrees of perceived threat from American Muslims and Muslim immigrants.

Integrated Threat Theory Stephan and Stephan’s (2000) ITT has been used to explain atti- tudes of both out-groups and in-groups and applied successfully in predicting the in-group’s attitude toward immigrants (Littleford, Wright, & Sayoc-Parial, 2005; Spencer-Rodgers & McGovern, 2002; Stephan, Diaz-Loving, & Duran, 2000; Stephan & Stephan, 1996, 2000). ITT identifies four types of threats, realistic threats, symbolic threats, intergroup anxiety, and negative stereotypes (Stephan & Stephan, 2000), and describes the antecedents as well as the outcomes of in-group–out-group communication.

Realistic threats are threats presented by the out-group and include threats to the in-group’s existence, political or economic power, and physical or material welfare. Symbolic threats are threats to the in-group’s worldview that involve norms, values, beliefs, at- titudes, and morals. Intergroup anxiety is the notion that people feel threatened during intergroup interactions. Negative stereo- types generate negative expectations, which, in turn, lead in-group members to anticipate negative consequences from intercultural encounters (Stephan & Stephan, 2000).

In a revised version of ITT, Stephan and Renfro (2002) posited that several antecedents are likely to increase perceived threat. Their

The Northwest Journal of Communication_Interior.indb 11 5/3/16 4:42 PM

12 KHAN AND BRUSCHKE

new antecedents are intergroup relations, cultural dimensions, in- dividual differences, and situational factors. Although Stephan and Renfro call for cross-cultural studies to test the impact of cultural dimensions on perceived threats, they refer specifically to three dimensions of culture: individualistic/collectivistic, high power distance/low power distance, and high uncertainty avoidance/low uncertainty avoidance.

We seek to supplement understanding of threat antecedents in two ways. First, it is possible that media images can have an impact on the formation of people’s levels of perceived threat and their ensuing attitudes and behaviors, thus making media exposure an additional antecedent. The media’s repetitive use of stereotypes, in particular, can play a key role in forming people’s attitudes toward out-group members. If contact between in-group and out-group members is rare, the only means by which in-group members can obtain information about the out-group is through secondhand accounts, and media images are likely to play a substantial role. This proposed function is tested with Hypothesis 1.

Second, ethnocentrism of in-group members may result in greater perceptions of threat and might be considered a specific individual difference, as suggested by Stephan and Renfro (2002). Although the concept has an extended lineage (Levinson, 1950; Sumner, 1906), a contemporary definition of ethnocentrism is of- fered by Neuliep and McCroskey (1997), who define ethnocentrism as a state where individuals see their cultural group as being the best one and reject other cultures as being inferior because of perceived differences or a disagreement between cultural ideologies. Although some degree of ethnocentrism is inevitable or even positive (Gu- dykunst & Nishida, 1994), people with ethnocentric attitudes engage in cooperative and obedient behaviors with in-group members but are competitive and disobedient with out-group members, hold be- littling attitudes toward them, and perform antisocial behaviors to- ward out-group members (Neuliep, Chaudoir, & McCroskey, 2001).

The Northwest Journal of Communication_Interior.indb 12 5/3/16 4:42 PM

13Media Coverage of Muslims

An application of ITT to American Muslims in a post–September 11 world suggests a second hypothesis:

h2: Greater degrees of ethnocentrism will be associated with a greater perceived threat from American Muslims and Muslim immigrants.

Intercultural Contact While some academicians argue that the clash of civilizations is an inevitable source of conflict and argue for separation (e.g., Huntington, 1993), the field of intercultural communication sees more and positive connections as the key to bridging gaps. The most developed academic treatment of the topic is the contact hypothesis, which posits that prejudice against members of one group by members of another will be reduced with increased social contact between members of the groups, especially if the contact is positive and individuals from different groups have the opportunity to interact on a personal level (Allport, 1954/1979). Intercultural contact research dates back to the 1950s, and a recent meta-analysis of intergroup contact theory has yielded valuable insights about the contact hypothesis (Allport, 1954/1979; Pettigrew & Tropp, 2000, 2006). Nearly 51% of all contact theory studies have focused on ethnic and racial groups, while the remaining studies have focused on target groups organized by sexual orientation, the elderly, the physically challenged, the mentally challenged, and mentally ill participants (Pettigrew & Tropp, 2006).

Reviews of the intergroup contact literature conclusively support the contact hypothesis, indicating that intergroup contact typi- cally reduces intergroup prejudice. Pettigrew and Tropp’s (2006) meta-analysis of 713 independent samples and 515 studies found overall support for the contact hypothesis, concluding that “mere” contact was sufficient to produce positive outcomes and that “there is little need to demonstrate further contact’s general ability to

The Northwest Journal of Communication_Interior.indb 13 5/3/16 4:42 PM

14 KHAN AND BRUSCHKE

lessen prejudice” (p. 768). Pettigrew and Tropp concluded that a more productive research direction is to explore factors that inhibit positive contact outcomes rather than factors that facilitate posi- tive contact outcomes: “Factors that curb contact’s ability to reduce prejudice are now the most problematic theoretically, yet the least understood” (p. 767). In particular, they believed perceived threat is such an inhibitor and encouraged research in this “fresh perspective” (p. 767). Pettigrew and Tropp ended with a call for models that consider “individual, structural, and normative antecedents of contact” (p. 768), such as the multilevel, multitheory framework utilized here. We propose adding ethnocentrism as a unique ante- cedent to ITT and connecting ITT to the contact hypothesis, which also has not studied ethnocentrism as an antecedent. Pettigrew and Tropp’s meta-analysis, for example, reviewed findings on prejudice, intergroup differentiation, in-group pride, and willingness to forgive an out-group but made no mention of ethnocentrism.

Knowledge is an additional factor related to contact, and nu- merous studies have reported that European Americans share a general lack of knowledge about cultures other than their own (see R. L. Jackson & Crawley, 2003, and references therein). This remains true despite an overall trend toward increased diversity (Ness & Kim, 2001). Theorizing causal directions is difficult, but a lack of knowledge about other cultures, and Muslim culture in particular, might also be a predictor of low intercultural contact.

A final reason to conduct more research on the contact hypothesis is the United States’s unique historical circumstance. The present study provides a timely and poignant test of the ability of simple contact to reduce negative attitudes about a group very likely to have generated anxiety in the years immediately following the 2001 attacks. A positive finding would confirm the ability of contact to reduce negative outcomes not simply in the context of general out- groups (the disabled, HIV/AIDS victims, those of other religions) but even in the context of a specific out-group aggressively portrayed

The Northwest Journal of Communication_Interior.indb 14 5/3/16 4:42 PM

15Media Coverage of Muslims

as a direct military threat. Given the sociopolitical ramifications of the War on Terror, we believe it is important to study the effects of the Bush administration’s rhetoric on the Muslim population in the United States to confirm whether contact theory can be extended to this particularly high-profile target group. Given the reasons to suspect that ethnocentrism and knowledge may be associated with contact, we predict the following:

h3: Ethnocentrism will be negatively associated with intercultural contact, and knowledge will be positively associated with inter- cultural contact.

MeTHods ANd resULTs

Respondents A total of 252 students from a major Southern California university volunteered to participate in the study. Of the 252 respondents, there were 94 (37%) men and 158 (63%) women. The average age for this study was 20.2 years. Respondents self-reported their eth- nicity, with 25% responding European American, 23% responding Asian American, 35% responding Latino American, 6% responding African American, 4% responding Middle Eastern American, and 8% responding other (Pacific Islander, Italian American, Native American, and of mixed ethnic background). Students were asked whether they had completed a general education course designed to increase appreciation for cultural diversity.

Initial Model Specification On the basis of prior research, the three theories were integrated into a single path model, as shown in Figure 1. Generally speaking, it is hypothesized that media variables will influence knowledge and attitudes and that knowledge and attitudes will subsequently predict the amount of contact with Muslims. The purpose of composing

The Northwest Journal of Communication_Interior.indb 15 5/3/16 4:42 PM

16 KHAN AND BRUSCHKE

the model this way is to study antecedents to contact with Muslims. Media variables appear at the top of the model. The three boxes

at the top left indicate gross amounts of media exposure. The three boxes at the top right indicate the degree of trust in media, self-reported media influence, and the interaction between trust in the media and total exposure. The interaction is based on the

figure 1. Full initial models with error weights set to 1 and all variables included.

The Northwest Journal of Communication_Interior.indb 16 5/3/16 4:42 PM

17Media Coverage of Muslims

supposition that those who trust the media the most will be influ- enced by exposure to the greatest extent. Boxes on the middle left include three measures of knowledge about Muslims, described later. Boxes on the middle right include two attitude measures: perceived threat from Muslims and ethnocentrism. The box at the bottom is contact with Muslims. All variables were manifest; no latent variables were explored.

Questionnaire and Measurement Although studies in cultivation theory have adopted different metrics, the core measure is one of total exposure. Other refined measures utilized in cultivation theory research include content- based measures, measures of news credibility and fear, and perceived television realism (Banjo, 2002; Gerbner & Gross, 1976). To assess gross exposure to media, three single-item measures were used. Participants self-reported the average number of hours per day they spent watching television, listening to the radio, and accessing the Internet. Three items measuring credibility were adapted from Gaziano and McGrath (1986), and three direct measures of self- perceived linkages between media and fear were included. The three media exposure items did not correlate, indicating that exposure is media specific (i.e., heavy television users are not necessarily heavy Internet users). The credibility items did not obtain accept- able reliability (α < .5), indicating that the items were measuring separate constructs. The item that most directly measured trust in media coverage of Muslims (“The media can be trusted with its coverage of Muslims”) was retained for analysis.

Measures for assessing perceived threat were adapted from Stephan and Stephan (1985, 1993), and all questions were made specific to Muslims. The seven items for the realistic threat measure focused on possible problems due to differences in values and beliefs between participants and foreign immigrants. An item from this scale is “Muslim immigration is undermining American culture.”

The Northwest Journal of Communication_Interior.indb 17 5/3/16 4:42 PM

18 KHAN AND BRUSCHKE

The eight items for the symbolic threat measure focused on crime, job loss, and economic costs due to immigration. A sample item is “Social services have become less available to Americans because of Muslim immigration.” Items for the intergroup anxiety measure fo- cused on people feeling threatened during intercultural interactions with Muslims because they were worried about negative outcomes. An item from this scale is “I am concerned about another terrorist attack carried out by Muslims.” Finally, the last component of ITT was negative stereotyping, and a measure of this construct focused on negative expectations of Muslims. An example of this scale is “Most Muslims are extremists.” The 20 items were combined into a single measure of perceived threat and were rated on a 5-point Likert-type scale ranging from 1 to 5, with a larger value indicating greater perceived threat. Of the 20 questions, 4 questions proved to be unreliable and were eliminated. The subsequent reliability for the perceived threat measure was a Cronbach’s alpha of .79.

A crucial antecedent for ITT is ignorance of the out-group. Knowledge about Muslims was measured in three ways. First, respondents were queried whether being Muslim was synony- mous with being from the Middle East, a common stereotype. An affirmative answer indicated ignorance about Muslims. Second, respondents were presented with a list of 10 paired names. Each name pair was similar, but only one of the names was Muslim; the number of correctly identified Muslim names was summed, with higher scores indicating greater knowledge about Muslims. Third, one of the discarded threat perception items queried whether there were multiple versions of Islam; higher scores indicated a greater knowledge about Muslims. Together, these three items served as measures of knowledge about Muslims.

To assess intercultural contact with Muslims, the participants responded to four items adapted from Biernat and Crandall (1996). The questions were as follows: (a) “How often do you talk to and engage in informal conversations with Muslims?” (b) “How often

The Northwest Journal of Communication_Interior.indb 18 5/3/16 4:42 PM

19Media Coverage of Muslims

do you study or do other class work with Muslims?” (c) “How often do you talk to and engage with Muslims at work?” and (d) “How often do you do things socially with Muslims (this includes things like sharing meals, going to movies and parties, etc.)?” The items were rated on a 5-point Likert-type scale ranging from 1 (never) to 5 (all the time). In addition, respondents were asked to indicate the number of Muslim friends they had. As a validation check, respondents were asked to list the names of their Muslim friends and, on a 1–4 scale, rate the amount of time they spent together. The reported number of Muslim friends, the average time spent with Muslim friends, and the greatest amount of time spent with any individual Muslim friend were combined with the four Biernat and Crandall items. The scale obtained a Cronbach’s alpha score of .85. All items were normalized with z-transformations and combined into a single index.

The ethnocentrism measure was adapted from Neuliep and McCroskey’s (1997) Generalized Ethnocentrism Scale. It consisted of 15 items and was adapted to assess evaluations of Muslims spe- cifically. An sample item is as follows: “Muslim culture would be smart to look up to our culture.” The items were rated on a 5-point Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree). The scale’s reliability was high (α = .87), and higher scores corresponded to greater ethnocentrism.

Descriptive Statistics Six variables were included in the final analysis: reported television exposure (M = 2.01, SD = 1.28), reported radio exposure (M = 2.29, SD = 2.25), reported Internet exposure (M = 2.87, SD = 1.93), per- ceived threat (M = 2.72, SD = .47), intercultural contact (M = 8.03, SD = 3.58), and ethnocentrism (M = 2.18, SD = .57). There was no significant difference on either ethnocentrism or perceived threat between those students who had completed the general education cultural diversity course and those who had not.

The Northwest Journal of Communication_Interior.indb 19 5/3/16 4:42 PM

20 KHAN AND BRUSCHKE

Model Testing and Development Because very few comprehensive models will obtain good mea- sures of data fit (Holbert & Stephenson, 2002, reported that 80% of published communication models do not obtain results indi- cating adequate data fit), a common approach is to test and then refine the model. This process, utilized here, is best described as an exploratory model construction.

Unsurprisingly, the model in Figure 1 did not obtain acceptable measures of data fit. All regression weights for endogenous vari- able error terms were set to 1, giving a standard regression output (Arbuckle, 2009). The minimum chi-square test (CMIN) was significant, χ2(2) = 678.7, p < .001, and the root mean square error adjusted (RMSEA) score was .30, above the accepted .05 criterion (Hu & Bentler, 1999, noted that the RMSEA criterion score is not a level of statistical significance). The model was subsequently modi- fied in three steps. First, all nonsignificant paths were eliminated. Second, the model was run again, and two of the remaining paths did not obtain significance in the new model. Those paths were subsequently eliminated. Third, a path between self-reported media influence and the belief that all Muslims were from the Middle East, while significant, was weak and was eliminated because it did not connect via any pathway to the ultimate dependent vari- able (contact).

The final model is displayed in Figure 2. All pathways in Figure 2 are statistically significant; numbers adjacent to pathway lines are beta weights, whereas figures to the top right of individual parameter boxes are r 2 scores. The final model did obtain adequate measures for goodness of fit. The CMIN chi-square was not significant, χ2(14) = 22.2, p = .07. The RMSEA score was .048, a marginal but acceptable figure. The model explained 20% of the variance in ethnocentrism, 42% of the variance in perceived threat, and 16% of the variance in contact.

The Northwest Journal of Communication_Interior.indb 20 5/3/16 4:42 PM

21Media Coverage of Muslims

dIsCUssIoN

While Stephan and Stephan’s (2000) ITT is most prevalently used to study prejudice and attitudes (Stephan et al., 2002; Stephan et al., 2000), this study combined ITT with cultivation theory to

figure 2. Final model with beta weights along pathways and r 2 scores for endogenous parameters.

The Northwest Journal of Communication_Interior.indb 21 5/3/16 4:42 PM

22 KHAN AND BRUSCHKE

explore the means by which individuals become aware of potential threats. Furthermore, this investigation also explored threat as an antecedent to contact, a key area of exploration identified by Stephan and Renfro (2002).

Cultivation Theory (Hypothesis 1) Total amount of media exposure did not correlate with any attitude in this study, nor was total exposure more likely to influence people who indicated a trust in the media. However, self-reported media influence (as indicated by agreement with the statement “After I watch the news/read news articles, I feel scared of Middle East- erners”) was related to both perceived threat and ethnocentrism, while trust in the media did predict ethnocentrism. Thus greater reported exposure to media coverage (via television, radio, and/or the Internet), by itself, was not associated with greater degrees of perceived threat. Possible explanations for these findings include (a) social desirability bias, (b) media saturation, (c) unreliability of measurement, and (d) the relative importance of content rather than sheer exposure.

A potential limitation of this study is that a social desirability bias may have been present. It is possible that the participants of this investigation may have falsely reported their levels of media exposure, perceived threat, or ethnocentrism because they were trying to preserve their self-esteem or portray themselves in a fa- vorable light to the researcher (Phillips & Clancy, 1972). However, if a social desirability effect was present, it contaminated only the relationships involving self-reports of media exposure, because intercultural contact proved to be systematically related to perceived threat and ethnocentrism. Although we suspect that some level of social desirability may have caused respondents to underreport levels of ethnocentrism and perceived threat, we do not believe that such an effect can explain the nonsignificance of Hypothesis 1. Also arguing against a social desirability effect was the lack of a

The Northwest Journal of Communication_Interior.indb 22 5/3/16 4:42 PM

23Media Coverage of Muslims

significant difference for either ethnocentrism or perceived threat between students who had and had not completed a course designed to enhance appreciation for cultural diversity. Normal guarantees of anonymity were offered on the questionnaire, further mitigating any social desirability bias.

A second possibility is that media saturation of the War on Terror was so complete that relative levels of media exposure are irrelevant. Simply put, it may be the case that coverage of Muslims was so ubiquitous and so negative that virtually everyone in the sample had received a level of exposure sufficient to shape attitudes about Muslims. In the aftermath of September 11, 2001, virtually all media were entirely consumed by the events. People watched more television and read more newspapers in the aftermath of the attacks (Cho et al., 2003; Scheufele et al., 2005). Increases in the use of all media might have made exposure patterns less distinguishable (Cho et al., 2003). Such a unique moment in media history could simply have created a situation of exposure so massive that cultivation theory did not anticipate it. It is quite possible that, regardless of the level of media exposure one experiences now, more than a decade after the attack, the level of exposure immediately after the attack was sufficient to affect most people in the sample. The significant pathway between trust in the media and ethnocentrism supports this interpretation; a correlation between the variables only seems possible if the overall media message propagates ethnocentric values (in this instance, the belief that Muslim culture is inferior to Western culture).

A third possible reason for the nonsignificance of Hypothesis 1 is a lack of reliability of the media exposure measures. The more sophisticated measures of media credibility utilized elsewhere did not produce reliable item clusters in the present experiment. The result was reliance on single-item measures of total exposure, which has inherent limitations (S. Jackson, 1992). This relatively poor measurement of the media exposure variable might be responsible

The Northwest Journal of Communication_Interior.indb 23 5/3/16 4:42 PM

24 KHAN AND BRUSCHKE

for the variable’s failure to covary with dependent measures. A fourth explanation concerns media content. This investigation

did not study media content. Perceived threat might depend on the type of media exposure, specifically news media versus other genres. For example, Meet the Press may have profoundly different content than the fictional Jack Bauer’s 24, and the pages of the New York Times may contain fundamentally different content than The Rush Limbaugh Show. Cultivation theory starts with the premise that content is so homogenous that the content itself is not as important as the volume of exposure. Future research may seek to explore the content of media as well as the volume of exposure.

A final possibility, of course, is that cultivation theory has simply been disconfirmed.

It is important to note that although raw media exposure did not emerge as a key model term, self-reported media influence did appear to play an important role, contributing both to perceived threat and ethnocentrism. Although both perceived threat and ethnocentrism were predicted through two paths, in each instance, self-reported media influence explained twice the variance as the other variable; that is, self-reported media influence explained roughly twice the variance of perceived threat than a belief in multiple versions of Islam did (the knowledge variable), and it explained roughly twice the variance in ethnocentrism that trust in the media did. The implication for cultivation theory is that media consumers are self-aware of the meanings they take from messages; in this study, the ability to indicate that “after I watch the news/read news articles, I feel scared of Middle Easterners” was significantly related to attitudes, whereas raw exposure measures were not. Evidently, self-awareness of the media interpretations seems to be important, at least in this context.

This finding contributes to the well-known discussion of the tension between cultivation theory’s view of audience members as fairly passive and uses and gratification theory’s view of audience

The Northwest Journal of Communication_Interior.indb 24 5/3/16 4:42 PM

25Media Coverage of Muslims

members as fairly active. At a minimum, the results obtained here suggest that it would be productive for cultivation theory to acknowledge a more conscious role of audience members. Results in the present study emerged only for those respondents who were self-aware that their exposure to news media was producing feel- ings of fear. We believe that this exploratory effort suggests this interesting pattern for future research: Specific media messages (in this instance, the Bush administration’s War on Terror) may so saturate a culture that relative levels of exposure are unimportant to attitude formation; however, individuals who can consciously identify that media messages make them afraid are more likely to develop negative attitudes.

Integrated Threat Theory (Hypothesis 2) As predicted, ethnocentrism was a significant predictor of perceived threat, although it did not directly serve as a disinhibitor for contact. Knowledge of other cultures did seem to play a role but was not an especially strong antecedent for perceived threat or contact. Belief in multiple versions of Islam did negatively correlate with perceived threat, but ethnocentrism explained 5 times more variance. Neither of the other two measures of knowledge correlated with perceived threat, although the ability to correctly identify Muslim names did correlate with contact (discussed in the next section).

This finding is an important contribution to ITT. As currently articulated, ITT supposes that some specific relationship between the in-group and out-group is necessary. For example, the in-group and out-group must have a certain status relation, or prior negative contact, or the in-group must be ignorant of the out-group. Our findings here suggest that, additionally, in-group ethnocentrism plays a role in threat perception, independent of any characteristic of the out-group or knowledge about them. The media findings do further suggest that information obtained from the media, at least for self-aware media consumers, does play a role in generating

The Northwest Journal of Communication_Interior.indb 25 5/3/16 4:42 PM

26 KHAN AND BRUSCHKE

perceptions of threat. This role does appear to be independent of actual knowledge about the out-group. In sum, ethnocentrism and awareness of media messages do appear to be important, and previously unexplored, antecedents to perceptions of threat.

Contact (Hypothesis 3) We discovered two direct predictors of contact: the ability to cor- rectly identify Muslim names (which was positively associated with contact) and perceived threat (which was negatively associated with contact). The inhibiting affect of perceived threat was roughly twice as large as the disinhibiting influence of correct name identification.

The influence of correct name recognition responds to Stephan and Renfro’s (2002) call for greater study of contact hypothesis antecedents by borrowing concepts from ITT. ITT suggests that one antecedent to threat perception is knowledge, and to the extent that correct name recognition is a measure of knowledge, it also appears that knowledge is an antecedent to contact. Interestingly, correct name recognition did not correlate with threat perception, and thus the relationship appears not to confirm ITT but does suggest a precursor for the contact hypothesis. Similarly, perceived threat, a concept central to ITT, does appear to function as a crucial antecedent to contact.

An inherent limitation of the correlation approach, not remedied by path modeling, is establishing causality. It may simply be that people who have contact with Muslims have greater knowledge about Muslim culture and fear them less, and such an interpreta- tion is entirely consistent with the data presented here. Regardless of direction, this investigation supports the extensive amount of work that has reported a negative relationship between intercultural contact, perceived threat, and ethnocentrism (Allport, 1954/1979; Brown & Hewstone, 2005; Hewstone, 2003; Pettigrew & Tropp, 2000; Stephan et al., 2002; Stephan et al., 2000; Voci & Hewstone, 2003; Ward & Masgoret, 2006).

The Northwest Journal of Communication_Interior.indb 26 5/3/16 4:42 PM

27Media Coverage of Muslims

We feel that it is significant to discover that the contact hypoth- esis does appear to hold in the context of American Muslims and Muslim immigrants to the United States even in the wake of the September 11 event and the subsequent barrage of negative media coverage of Muslims. Allport’s (1954/1979) conditions of equal status, common goals, intergroup cooperation, and authority sanction have informed the literature on intergroup relations. Pettigrew (1998) added friendship potential as a fifth condition for positive intergroup contact. Intercultural interactions with high friendship potential allow participants to develop a greater intellectual and relational capacity because of their meaningful, reciprocal exchange (Nagda & Zúñiga, 2003). Studies have shown the importance of intergroup friendship in promoting positive changes in intergroup attitudes (Aboud, Mendelson, & Purdy, 2003; Pettigrew, 1997; Wright, Aron, & Tropp, 2002; Wright, Brody, & Aron, 2005; Wright & Tropp, 2005). This study extends and highlights the importance of friendship with culturally different others and the significant role that intercultural contact plays in reducing the effects of perceived threat on ethnocentrism. It is a sharp rejection of approaches, such as those offered by Huntington (1993), that recommend a decrease in personal contact as a means to safety.

LIMITATIONS

Although consistent with and derived from prior research, the media use measures adopted here do seem problematic. Measures of raw media consumption were single-item measures; measures of media credibility that were reliable in prior research did not form scales here and forced the use of single-item measures of media credibility as well. Although we do not wish to minimize the limitations of single-item measures (S. Jackson, 1992), the perfor- mance of the self-awareness of media influence measure item does suggest some substantive relationship at play. Nonetheless, future

The Northwest Journal of Communication_Interior.indb 27 5/3/16 4:42 PM

28 KHAN AND BRUSCHKE

research should seek more refined measures of media consumption and media consumer awareness. The methods here also relied on respondent self-reporting and discussed total media usage; future research might explore distinctions between news and entertain- ment media or query attention to messages particular to national security or depictions of Muslims.

Reliability limitations do seem limited to the media measures; contact, ethnocentrism, and perceived threat scales all emerged with acceptable reliability scores.

Although this study did explore potential additional anteced- ents to ITT, any comparison to Stephan and Stephan’s (1990) four extant factors was not possible; only knowledge was explored in any way. Future research should include the additional three fac- tors as well. It should be noted that the knowledge measures used here did not correlate with one another. It is unclear whether the measures therefore indicate separate dimensions of knowledge or different rigor in the respondent task (correctly identifying names as Muslim versus providing Likert-style answers) or whether they are simply not reliable. Future research should seek to measure cultural knowledge in more exacting terms.

Finally, the use of correlation data makes any causal claim dif- ficult to sustain (although it would not be surprising to discover some mutually reinforcing influence in force). Untangling the direction of perceived threat, knowledge, and contact relationships is important work for future research.

CoNCLUsIoNs

The hope at the outset of this study was that combining the three theories in a complementary way would provide a more complete account of one of the more urgent practical situations U.S. society faces and might also offer the opportunity for a reflective reformula- tion of theory. We hoped to combine cultivation theory with ITT

The Northwest Journal of Communication_Interior.indb 28 5/3/16 4:42 PM

29Media Coverage of Muslims

to enrich understandings of antecedents to intercultural contact. The various parameters from these theories were incorporated into an exploratory path model.

The result was partially successful. The cultivation theory ex- pectation that media exposure would result in perceived threat did not materialize, perhaps due to poor measurement, but more pro- vocatively perhaps due to media saturation. However, respondents who reported a greater degree of self-aware media influence were more likely to be both ethnocentric and more likely to perceive a threat from Muslims. This suggests a direction for cultivation theory that increasingly incorporates the influence of audience engagement into its explanations. Furthermore, cultivation theory would seem to benefit from an accounting of situations of extreme media saturation.

The predicted relationship between perceived threat and ethno- centrism was confirmed, extending ITT in an important applied context. To the extent that minority cultural members are perceived as a threat, it may be expected that they will be viewed by majority cultural members in negative ways. In addition, characteristics of the in-group, such as ethnocentrism, may play an active role in the degree to which out-groups are perceived as threats. This contribution is both theoretically important and practically mean- ingful; ethnocentrism was the largest single predictor of perceived threat in a model that explained 42% of the variance in perceived threat. It does appear that ethnocentrism is a concept that should be incorporated into ITT.

Perceived threat was a significant antecedent to contact with Muslims. Unexpectedly, knowledge of out-group culture emerged as a direct predictor of contact, even though it did not correlate with perceived threat, as expected. Together, these two factors accounted for 16% of the variance in contact. On one hand, the relationships uncovered here do make a contribution to understanding the forces that encourage and discourage contact between in-groups

The Northwest Journal of Communication_Interior.indb 29 5/3/16 4:42 PM

30 KHAN AND BRUSCHKE

and out-groups. On the other hand, there is clearly much more that can be understood about the antecedents of in-group contact and much ground for future research to cover. The exploratory model offered here can serve as a useful starting point for more extensive and confirmatory research.

Although questions of causality remain elusive, this investiga- tion concludes that the solution to cultural conflict is an increase, and not a decrease, in intercultural contact. The study of ways to improve intercultural contact seems more urgent than ever in our current political context.

reFereNCes

Aboud, F. E., Mendelson, M. J., & Purdy, K. T. (2003). Cross-race peer relations and friendship quality. International Journal of Behavioral Development, 27, 165–173.

Akram, S. M. (2002). The aftermath of September 11, 2001: The targeting of Arabs and Muslims in America. Arab Studies Quarterly, 24, 61–118.

Allport, G. (1979). The nature of prejudice. Cambridge, MA: Addison- Wesley. (Original work published 1954)

Arbuckle, J. L. (2009). Amos 18 user’s guide. Chicago: SPSS Inc. Argyrides, M., & Downey, J. L. (2004). September 11: Immediate and

long term effects on measures of aggression, prejudice, and person perception. North American Journal of Psychology, 6, 175–188.

Banjo, O. (2002). The effect of media consumption on the perception of romantic relationships. Penn State McNair Journal, 9, 9–33.

Biernat, M., & Crandall, C. (1996). Creating stereotypes and capturing their content. European Journal of Social Psychology, 26, 867–898.

Brown, R., & Hewstone, M. (2005). An integrative theory of intergroup contact. In M. P. Zanna (Ed.), Advances in experimental social psy- chology (Vol. 37, pp. 255–343). San Diego, CA: Elsevier.

Bryant, J., Carveth, R. A., & Brown, D. (1981). Television viewing and anxiety: An experimental investigation. Journal of Communication, 31, 106–119.

Cho, J., Boyle, M. P., Keum, H., Shevy, M. D., McLeod, D. M., Shah, D. V., & Pan, Z. (2003). Media, terrorism, and emotionality: Emotional

The Northwest Journal of Communication_Interior.indb 30 5/3/16 4:42 PM

31Media Coverage of Muslims

differences in media content and public reactions to the September 11th terrorist attacks. Journal of Broadcasting and Electronic Media, 47(3), 309–327.

Elasmar, M. G. (2008). Through their eyes: Factors affecting Muslim sup- port for the U.S.-led War on Terror. Spokane, WA: Marquette Books.

Gaziano, C., & McGrath, K. (1986). News credibility scale. In R. B. Ru- bin, P. Palmgreen, & H. E. Sypher (Eds.), Communication research measures: A sourcebook (pp. 234–237). New York: Guilford Press.

Gerbner, G., & Gross, L. (1976). Living with television: The violence profile. Journal of Communication, 26, 172–199.

Gerbner, G., Gross, L., Morgan, M., & Signorielli, N. (1994). Grow- ing up with television: The cultivation perspective. In J. Bryant & D. Zillmann (Eds.), Media effects: Advances in theory and research (pp. 17–41). Hillsdale, NJ: Lawrence Erlbaum Associates.

Gudykunst, W. B., & Nishida, T. (1994). Bridging Japanese/North American differences. Thousand Oaks, CA: Sage.

Hewstone, M. (2003). Intergroup contact: Panacea for prejudice? The Psychologist, 16, 352–355.

Holbert, R. L., & Stephenson, M. T. (2002). Structural equation modeling in the communication sciences, 1995–2000. Human Communication Research, 28, 531–551.

Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55.

Huntington, S. P. (1993). The clash of civilizations? Foreign Affairs, 72, 22–49. Jackson, R. L., & Crawley, R. L. (2003). White student confessions about

a Black male professor: A cultural contracts theory approach to intimate conversations about race and worldview. Journal of Men’s Studies, 12, 25–37.

Jackson, S. (1992). Message effects research: Principles of design and analysis. New York: Guilford Press.

Kellner, D. (2002). September 11, the media, and war fever. Television and New Media, 3, 143–151.

Levinson, D. J. (1950). Politico-economic ideology and group memberships in relation to ethnocentrism. In T. W. Adorno, E. Frenkel-Brunswik, D. J. Levinson, & R. N. Sanford (Eds.), The authoritarian personality (pp. 151–221). New York: Harper.

Littleford, L. N., Wright, M. O., & Sayoc-Parial, M. (2005). White students’ intergroup anxiety during same-race and interracial interactions: A multimethod approach. Basic and Applied Social Psychology, 27(1), 85–94.

The Northwest Journal of Communication_Interior.indb 31 5/3/16 4:42 PM

32 KHAN AND BRUSCHKE

Marcus, G. E., Sullivan, J. L., Theiss-Morse, E., & Wood, S. L. (1995). With malice toward some: How people make civil liberties judgments. New York: Cambridge University Press.

Media has anti-Muslim bias, claims report. (2005, November 14). The Guardian. Retrieved from http://media.guardian.co.uk/broadcast /story/0,7493,1642179,00.html

Nagda, B. R. A., & Zúñiga, X. (2003). Fostering meaningful racial engage- ment through intergroup dialogues. Group Processes and Intergroup Relations, 6, 111–128.

Ness, C., & Kim, R. (2001, March 13). Census reveals fast-growing diversity in US. San Francisco Chronicle, p. B2.

Neuliep, J. W., Chaudoir, M., & McCroskey, J. C. (2001). A cross-cultural comparison of ethnocentrism among Japanese and United States college students. Communication Research Reports, 18(2), 137–146.

Neuliep, J. W., & McCroskey, J. C. (1997). The development of a U.S. and generalized ethnocentrism scale. Communication Research Reports, 14(4), 385–398.

Nisbet, E. C., Ostman, R., & Shanahan, J. (2009). Public opinion toward Muslim Americans: Civil liberties and the role of religiosity, ideol- ogy, and media use. In A. Sinno (Ed.), Muslims in Western politics (pp. 161–199). Bloomington: Indiana University Press.

Pettigrew, T. F. (1997). Generalized intergroup contact effects on prejudice. Personality and Social Psychology Bulletin, 23, 173–185.

Pettigrew, T. F. (1998). Intergroup contact theory. Annual Review of Psy- chology, 49, 65–85.

Pettigrew, T. F., & Tropp, L. (2000). Does intergroup contact reduce preju- dice? Recent meta-analytic findings. In S. Oskamp (Ed.), Reducing prejudice and discrimination (pp. 93–114). Mahwah, NJ: Lawrence Erlbaum Associates.

Pettigrew, T. F., & Tropp, L. R. (2006). A meta-analytic test of intergroup contact theory. Journal of Personality and Social Psychology, 90, 751–783.

Phillips, D. L., & Clancy, K. J. (1972). Some effects of social desirability in survey studies. American Journal of Sociology, 77, 921–938.

Pintak, L. (2006). Reflections in a bloodshot lens. London: Pluto Press. Prengaman, P. (2006, March 28). Poll shows growing fear among im-

migrants. New America Media. Retrieved from http://news.new americamedia.org/news/view_article.html?article_id=4bb26e36d97e3 007bdac122315159071

Romer, D., Jamieson, K. H., & Aday, S. (2003). Television news and the cultivation of fear of crime. Journal of Communication, 53, 88–104.

Said, E. W. (1979). Orientalism. New York: Vintage Books.

The Northwest Journal of Communication_Interior.indb 32 5/3/16 4:42 PM

33Media Coverage of Muslims

Scheufele, D.  A., Nisbet, M.  C., & Ostman, R. (2005). September 11th news coverage, public opinion, and support for civil liberties. Mass Communication and Society, 8(3), 197–218.

Shaheen, J.  G. (2001). Reel bad Arabs: How Hollywood vilifies a people. Northampton, MA: Interlink.

Sheridan, L. P., & Gillett, R. (2005). Short note: Major world events and discrimination. Asian Journal of Social Psychology, 8, 191–197.

Spencer-Rodgers, J., & McGovern, T. (2002). Attitudes toward the culturally different: The role of intercultural communication barriers, affective responses, consensual stereotypes, and perceived threat. International Journal of Intercultural Relations, 26, 609–631.

Stephan, C. W., & Stephan, W. G. (1990). Two social psychologies (2nd ed.). Belmont, CA: Wadsworth.

Stephan, W. G., Boniecki, K. A., Ybarra, O., Bettencourt, A., Ervin, K. S., Jackson, L. A., . . . Renfro, C. L. (2002). The role of threats in the racial attitudes of Blacks and Whites. Personality and Social Psychology Bulletin, 28(9), 1242–1254.

Stephan, W. G., Diaz-Loving, R., & Duran, A. (2000). Integrated threat theory and intercultural attitudes: Mexico and the United States. Journal of Cross-Cultural Psychology, 31(2), 240–249.

Stephan, W. G., & Renfro, C. L. (2002). The role of threats in intergroup relations. In D. Mackie & E.  R. Smith (Eds.), From prejudice to intergroup emotions (pp. 191–208). New York: Psychology Press.

Stephan, W. G., & Stephan, C. W. (1985). Intergroup anxiety. Journal of Social Issues, 41, 157–176.

Stephan, W. G., & Stephan, C. W. (1993). Cognition and affect in stereotyp- ing: Parallel interactive networks. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 111–136). Orlando, FL: Academic.

Stephan, W. G., & Stephan, C. W. (1996). Predicting prejudice. International Journal of Intercultural Relations, 20(3/4), 409–426.

Stephan, W. G., & Stephan, C. W. (2000). An integrated threat theory of prejudice. In S. Oskamp (Ed.), Reducing prejudice and discrimination (pp. 23–46). Mahwah, NJ: Lawrence Erlbaum Associates.

Sumner, W. G. (1906). Folkways. Boston: Ginn. Sunstein, C. (2003). Terrorism and probability neglect. Journal of Risk

and Uncertainty, 26(2/3), 121–136. Sunstein, C. (2004). Fear and liberty. Social Research, 71(4), 976–996. Voci, A., & Hewstone, M. (2003). Intergroup contact and prejudice toward

immigrants in Italy: The mediational role of anxiety and the moderati- onal role of salience. Group Processes and Intergroup Relations, 6, 37–54.

The Northwest Journal of Communication_Interior.indb 33 5/3/16 4:42 PM

34 KHAN AND BRUSCHKE

Ward, C., & Masgoret, A. (2006). An integrative model of attitudes toward immigrants. International Journal of Intercultural Relations, 30, 671–682.

Western media coverage of Islam a failure: US expert. (2005, August 31). Islam Online. Retrieved from http://www.islamonline.net/English /News/2005-08/31/article06.shtml

Woods, J. (2007). What we talk about when we talk about terrorism: Elite press coverage of terrorism risk from 1997 to 2005. Harvard International Journal of Press/Politics, 12(3), 3–20.

Wright, S. C., Aron, A., & Tropp, L. R. (2002). Including others (and groups) in the self: Self-expansion and intergroup relations. In J. P. Forgas & K. D. Williams (Eds.), The social self: Cognitive, interpersonal, and intergroup perspectives (pp. 343–363). Philadelphia: Psychology Press.

Wright, S. C., Brody, S. A., & Aron, A. (2005). Intergroup contact: Still our best hope for improving intergroup relations. In C. Crandall & M. Schaller (Eds.), Social psychology of prejudice: Historical perspectives (pp. 115–142). Seattle, WA: Lewinian Press.

Wright, S. C., & Tropp, L. R. (2005). Language and intergroup contact: In- vestigating the impact of bilingual instruction on children’s intergroup attitudes. Group Processes and Intergroup Relations, 8(3), 309–328.

The Northwest Journal of Communication_Interior.indb 34 5/3/16 4:42 PM

Copyright of Northwest Journal of Communication is the property of Northwest Communication Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.