Reliance on Direct and Mediated Contact and Public Policies Supporting Outgroup Harm: Reliance on Direct and Mediated Contact

Abstract

Two studies examined the effects of reliance on direct and media-based contact for information about Muslims on Americans' stereotypic beliefs of and negative emotions toward Muslims and support for public policies harming Muslims domestically and internationally. Results revealed that reliance on media for information about Muslims was positively associated with stereotypic beliefs, negative emotions, and support for harmful policies. Reliance on direct contact for information about Muslims produced the opposite results. Results from a three-wave longitudinal design revealed that reliance on media and direct contact significantly predict changes in negative emotions which then predict changes in support for civil restrictions for Muslim Americans. We discuss the differential effects of reliance on media-based and direct contact in influencing intergroup outcomes.

Full text

Journal of Communication ISSN 0021-9916
ORIGINAL ARTICLE
Reliance on Direct and Mediated Contact and
Public Policies Supporting Outgroup Harm
Muniba Saleem1, Grace S. Yang2, & Srividya Ramasubramanian3
1 Department of Communications & Research Center for Group Dynamics, Institute for Social Research, The
University of Michigan, Ann Arbor, MI 48109, USA
2 Department of Psychology & Research Center for Group Dynamics, Institute for Social Research, The
University of Michigan, Ann Arbor, MI 48109, USA
3 Department of Communication, Texas A&M University, College Station, TX 77843, USA
Two studies examined the eects of reliance on direct and media-based contact for infor-
mation about Muslims on Americans’ stereotypic beliefs of and negative emotions toward
Muslims and support for public policies harming Muslims domestically and internation-
ally. Results revealed that reliance on media for information about Muslims was positively
associated with stereotypic beliefs, negative emotions, and support for harmful policies.
Reliance on direct contact for information about Muslims produced the opposite results.
Results from a three-wave longitudinal design revealed that reliance on media and direct
contact signicantly predict changes in negative emotions which then predict changes in
support for civil restrictions for Muslim Americans. We discuss the dierential eects of
reliance on media-based and direct contact in inuencing intergroup outcomes.
Keywords: Contact, Media Eects, Prejudice, Intergroup, Public Policy.
doi:10.1111/jcom.12234
Extant research suggests that our understanding of outgroups is inuenced by vari-
ous socializing agents including family, peers, and media. Although direct forms of
intergroup contact in the United States have increased over the years due to changing
racial demographics and policies, many individuals still rely on media as their pri-
mary source of information about outgroups (Mutz & Goldman, 2010). For example,
themajorityofAmericansreportthatmediaistheirprimarysourceofinformation
about Muslims (Nisbet, Ostman, & Shanahan, 2009) and some studies reveal that the
inuence of media on Americans’ attitudes toward Muslims is stronger than that of
other informational sources (Kalkan, Layman, & Uslaner, 2009).
Past research suggests that the ways in which outgroups are represented in media
ultimately inuence people’s perceptions, attitudes, and behaviors toward members of
Corresponding author: Muniba Saleem; e-mail: saleemm@umich.edu
Journal of Communication (2016) © 2016 International Communication Association 1

Reliance on Direct and Mediated Contact M. Saleem et al.
those groups (e.g., Harwood, Hewstone, Amichai-Hamburger, & Tausch, 2012; Mas-
tro, 2009). More specically, media stereotypes of outgroups can inuence negative
attitudes and behaviors, including public policy decisions that harm members of the
depicted outgroup (e.g., Mastro & Kopacz, 2006; Ramasubramanian, 2011; Saleem,
Prot, Anderson, & Lemieux, 2015; Tan, Fujioka, & Tan, 2000). Although many minor-
ity groups are negatively depicted in American media, negative stereotypes of Mus-
lims are especially pervasive in media (Tukachinsky, Mastro, & Yarchi, 2015). Mus-
lims are frequently linked with violence, terrorism, and aggression across American
media outlets, including cable news (Dixon & Williams, 2015), newspapers (Nacos
& Torres-Reyna, 2007), and television and movies (Alsultany, 2012; Shaheen, 2009).
ese negative media portrayals inuence negative beliefs toward Arabs, Muslims,
and people of Middle Eastern descent (Das, Bushman, Bezemer, Kerkhof, & Ver-
meulen, 2009; Nisbet et al., 2009; Saleem & Anderson, 2013). Although these iden-
tities are distinct, Americans oen conate them as one identity and their attitudes
toward these dierent groups are remarkably similar (Kteily, Bruneau, Waytz, & Cot-
terill, 2015).
We examine the extent to which reliance on direct and media-based contact for
information about Muslims inuences Americans’ stereotypes, emotions, and sup-
port for public policies that harm Muslims using cross-sectional correlational and
longitudinal designs. is study extends previous research in at least six ways. First,
it is one of the few studies to simultaneously test the eects of reliance on direct and
media-based contact on outgroup attitudes and public policies. Second, it examines
the mediating roles of beliefs and emotions in understanding the link between reliance
on dierent sources for information about outgroups and support for public poli-
cies harming members of those groups. ird, it goes beyond assessing preferential
treatment of ingroup over outgroup members (i.e., ingroup positivity) and examines
support for public policies that represent outgroup harm (discussed in detail in sub-
sequent sections). Fourth, few empirical studies to date have examined the extent to
which media stereotypes of Muslims inuence Americans’ attitudes toward Muslims,
despite the prevalence of these stereotypes in American media. is study addresses
this important gap. Fih, the majority of past research on media stereotypes and pub-
lic policy is done using correlational designs, which limits the causal interpretation of
the results. We used a three-wave longitudinal design in Study 1 to provide stronger
evidence for the direction and causality of the hypothesized relationships. Finally,
Study 2 used a cross-sectional correlational design to generalize the ndings of Study
1 using a more representative adult sample. We borrow from intergroup contact the-
ory (Allport, 1954) and social cognitive theory (Bandura, 1986) to understand how
reliance on direct and media-based contact inuences intergroup outcomes.
Intergroup contact theory
Intergroupcontacttheorysuggeststhatcontactbetweenmembersofdierentgroups
can reduce prejudice and improve intergroup relations (Allport, 1954). Allport
2Journal of Communication (2016) © 2016 International Communication Association

M. Saleem et al. Reliance on Direct and Mediated Contact
theorized that contact under certain conditions would be optimal: equal status
between groups, institutional support of contact, shared goals between the groups
involved, and cooperation between groups to meet those goals. ough contact eects
under these conditions are the most powerful, contact in nonoptimal conditions can
also inuence positive intergroup outcomes (Pettigrew & Tropp, 2006). e majority
of work on intergroup contact has focused on direct (i.e., face-to-face) interactions
with outgroup members; however, recent work reveals that vicarious contact, includ-
ing media-based contact, can also inuence attitudes toward outgroups (Harwood
et al., 2012).
Media-based contact
e importance of observing and learning from others is well established within the
social cognitive theory framework (Bandura, 1986, 2001). Specically, real or virtual
observations inform our schemas and associated cognitions about the social world,
especially when they involve individuals with whom we identify (Bandura, 2001).
ese observations allow individuals to acquire new information or modify existing
schemas involving outgroups ultimately increasing the likelihood of modeling the
observed behavior in the short and long term (Bandura, 1986, 2001). Indeed, research
suggeststhatobservingpositiveintergroupcontact,includingmedia-baseddepic-
tions, can inuence positive attitudes toward outgroups (Joyce & Harwood, 2012;
Ortiz & Harwood, 2007; Schiappa, Gregg, & Hewes, 2005). Importantly, research
suggests that positive exemplars of outgroup members, even in the absence of ingroup
members,caninuencepositiveintergroupoutcomes.Forexample,media-based
exposure to exemplary outgroup members can increase: (a) perceptions that the
outgroup has been the victim of discrimination (Bodenhausen, Schwarz, Bless,
& Wänke, 1995); (b) external attributions of negative outgroup behavior (Power,
Murphy, & Coover, 1996); and (c) support for public policies that benet outgroup
members (Ramasubramanian, 2011).
Although encouraging, the fact that minorities are mostly under- or negatively
represented in American media (Behm-Morawitz & Ortiz, 2012; Tukachinsky et al.,
2015) dampens the enthusiasm for these positive eects. Indeed, numerous stud-
ies reveal that negative media stereotypes of outgroups inuence negative attitudes
and behavioral tendencies toward those groups (Mastro, 2009). ese ndings high-
light a major barrier in expecting media-based contact to have positive eects on
attitudes —if media portrayals of social groups are largely negative, then reliance on
media as a source of information about these groups will result in negative attitudes
and action tendencies toward that outgroup (Harwood et al., 2012).
Direct contact
Research examining face to face contact reveals that direct contact with outgroup
members can reduce negative beliefs, aect, and behaviors toward outgroup mem-
bers (Pettigrew & Tropp, 2006). Furthermore, high-quality contact (e.g., cross-group
friendships) is especially powerful in improving intergroup relations (Pettigrew &
Journal of Communication (2016) © 2016 International Communication Association 3

Reliance on Direct and Mediated Contact M. Saleem et al.
Tropp, 2006). However, negative direct contact, similar to negative media-based con-
tact, with outgroup members can result in deleterious outcomes. In fact, negative con-
tact can increase group salience resulting in stronger eects on prejudice compared to
positive contact (Barlow et al., 2012; Paolini, Harwood, & Rubin, 2010). is is con-
sistent with research revealing that people weigh negative information more heavily
than positive information (Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001).
Dierential eects of media-based and direct contact
e majority of studies have examined the eects of dierent types of contact in isola-
tion (cf., Mastro & Tropp, 2004; Ortiz & Harwood, 2007; Paolini et al., 1989; Schiappa
et al., 2005). is is problematic as individuals are likely exposed to and rely on both
forms of contact involving outgroups (Mutz & Goldman, 2010). us, it is imperative
to study the eects of dierent types of contact on intergroup outcomes as well as their
underlying mediating mechanisms (Harwood et al., 2012). Indeed, media stereotypes
are especially inuential when direct contact with outgroup members is limited (e.g.,
Armstrong, Neuendorf, & Brentar, 1992; Fujioka, 1999). Direct, especially positive,
contact with outgroup members provides individuals with new or alternative infor-
mation about an outgroup which can counteract any negative cognitions activated by
media stereotypes (Joyce & Harwood, 2012). Void of these positive experiences, media
stereotypes are likely to create and strengthen stereotypic cognitions of outgroups,
which over time become automatized and applied when interacting with outgroup
members (Bandura, 2001).
Reliance on media-based and direct contact for information about outgroups
Although exposure to and reliance on information are conceptually distinct, they
share some theoretical similarities in the context of intergroup dynamics. Media
dependency theory suggests that people are especially likely to rely on media for
information when their personal experience with or knowledge of the issue is
limited (DeFleur & Ball-Rokeach, 1989). is is consistent with contact theory
which suggests that media-based contact eects are especially inuential when direct
contact is limited (Harwood et al., 2012). Although contact theory does not explicitly
conceptualize exposure to dierent forms of contact as suggesting relative dierences
in the extent to which individuals rely on these contact sources for information,
many media eects scholars interpret these eects as such (e.g., Behm-Morawitz &
Ortiz, 2012; Mastro, Behm-Morawitz, & Ortiz, 2007; Ramasubramanian, 2013). For
example, Ramasubramanian (2013) discovered that Whites who relied on media
as opposed to direct contact as the primary source of information about African
Americans were more likely to express stereotypic beliefs and prejudice against
African Americans. In this study, participants had to report their primary sources
of information about African Americans as either direct contact or media. Rather
than examining the relative dierence of reliance on personal versus media-based
contact, a more interesting question is the extent to which reliance on direct and
media-based contact dierentially inuences attitudes and behaviors toward an
4Journal of Communication (2016) © 2016 International Communication Association

M. Saleem et al. Reliance on Direct and Mediated Contact
outgroup. We examine this question in this study with respect to support for public
policies harming Muslims.
Media and support for public policies harming outgroups
Media’s inuence on public policies is signicant as the majority of Americans
obtain their public aairs information primarily from the media (Delli Carpini &
Keeter, 1996), and media can have a disproportionate inuence on public opinion
(Ball-Rokeach & DeFleur, 1976). Indeed, correlational evidence suggests that Whites’
recollection of negative media portrayals of African Americans is signicantly associ-
ated with negative stereotypes of African Americans and in turn reduced support for
armative action policies (Mastro & Kopacz, 2006; Tan et al., 2000). Ramasubrama-
nian (2011) used an experimental design to show that White participants primed with
stereotypic images of African American celebrities, relative to participants primed
with counterstereotypic images, displayed more stereotypic beliefs and prejudicial
feelings, and were less likely to support armative action policies.
Note that all of these studies examine support for one specic public policy (i.e.,
armative action). Research reveals that support for armative action is motivated
by ingroup positivity (Lowery, Unzueta, Knowles, & Go, 2006), which is theoreti-
cally and conceptually distinct from intentions to harm an outgroup (Brewer, 2001,
2010). Ingroup positivity may reect positive sentiments (trust, empathy, and coop-
eration) toward the ingroup that are withheld from an outgroup. Outgroup harm,
however, entails active hostility, derogation, and intent to harm an outgroup without
any necessary benet to the ingroup (Brewer, 2001, 2010). Most forms of intergroup
bias occur with the primary motivation to benet the ingroup rather than harm the
outgroup (Mummendey & Otten, 2001).
However, these constraints disappear when outgroups are perceived as threaten-
ing and aggressive (Stephan & Stephan, 2000), or when outgroups are viewed with
hatred or contempt— emotions that justify outgroup harm above and beyond ingroup
benet (Mackie, Devos, & Smith, 2000). reatening outgroups arouse aggressive
cognitions, negative emotions, and behaviors representative of outgroup harm (Riek,
Mania, & Gaertner, 2006). Scholars further suggest that negative emotions toward an
outgroup are key in dierentiating motivations of ingroup love from outgroup harm
(Brewer, 2010; Mackie et al., 2000; Mummendey & Otten, 2001). is is important as
past studies on media stereotypes have mostly focused on outgroup cognitions rather
than emotions (cf., Atwell Seate & Mastro, 2015). We address this limitation in this
work by examining the mediating role of emotions and perceptions in understand-
ing the eects of reliance on direct and media-based contact on support for policies
harming Muslims.
Overview of the current studies
e above review of existing theory and empirical evidence suggests that Americans’
reliance on media for information about Muslims will positively inuence negative
Journal of Communication (2016) © 2016 International Communication Association 5

Reliance on Direct and Mediated Contact M. Saleem et al.
perceptions of, and emotions toward, Muslims, and subsequently support for public
policies harming Muslims. However, Americans’ reliance on direct contact, especially
their Muslim friends, for information about Muslims will inversely inuence these
outcomes.Wetestedthesepredictionsacrosstwostudiesusingtwodierentsamples
and designs.
Study 1
Study 1 used a three-wave longitudinal design to test the eects of reliance on direct
and media-based contact for information about Muslims on changes in perception
of, and emotions toward, Muslims, and support for policies harming Muslims. Based
on intergroup contact theory and media dependency theory, we expected reliance
on direct contact for information about Muslims at Time 1 to be negatively associ-
ated with perceptions of Muslims as aggressive and with negative emotions toward
Muslims at Time 2. Although contact in general is known to reduce prejudice, our
assessment of contact specically included friendships with Muslims, which is espe-
cially likely to inuence positive intergroup outcomes (Turner, Hewstone, & Voci,
2007). However, given that Muslims are represented negatively in American media
(see citations above), we expected reliance on media for information about Muslims
at Time 1 to be positively associated with perceptions of Muslims as aggressive and
with negative emotions toward Muslims at Time 2. Finally, consistent with previous
work, we expected an interaction between reliance on media and direct contact, such
that the eects of reliance on media on perceptions of, and emotions toward, Mus-
lims are signicant for those low, relative to high, on reliance on direct contact for
information about Muslims.
Hypotheses 1a and 1b: Reliance on direct contact as a source of information about Muslims
at Time 1 will be inversely associated with perceptions of Muslims as aggressive at Time 2
(H1a) and with negative emotions towards Muslims at Time 2 (H1b).
Hypotheses 2a and 2b: Reliance on media as a source of information about Muslims at
Time 1 will be positively associated with perceptions of Muslims as aggressive at Time 2
(H2a) and with negative emotions towards Muslims at Time 2 (H2b).
Hypotheses 3a and 3b: Reliance on media as a source of information about Muslims at
Time 1 will be positively associated with perceptions of Muslims as aggressive at Time 2 and
with negative emotions towards Muslims at Time 2 for those low, relative to high, on
reliance on direct contact for information about Muslims (H3a and H3b).
Based on past research predicting outgroup harm, we expected perceptions of
Muslims as aggressive and negative emotions towards Muslims to positively predict
support for public policies that harm Muslims internationally and domestically.
ough there is some evidence that thinking of outgroups within a superordinate
identity (e.g., American) reduces outgroup prejudice (Gaertner & Dovidio, 2000),
research suggests that Americans’ attitudes towards Muslim Americans are not
signicantly dierent than their attitudes towards Muslims living abroad (Kteily
et al., 2015; Saleem, Prot, Anderson, et al., 2015; Saleem, Prot, Cikara, et al., 2015;
6Journal of Communication (2016) © 2016 International Communication Association

M. Saleem et al. Reliance on Direct and Mediated Contact
Sides & Gross, 2013). us, we expected our hypothesized mediators (perceptions
and emotions) to inuence Americans’ support for public policies involving both
Muslims living abroad and Muslim Americans.
Hypotheses4aand4b:PerceptionsthatMuslimsareaggressiveatTime2willbepositively
associated with support for civil restrictions for Muslim Americans at Time 3 (H4a) and
with support for military action in Muslim countries at Time 3 (H4b).
Hypotheses 5a and 5b: Negative emotions toward Muslims at Time 2 will be positively
associated with support for civil restrictions for Muslim Americans at Time 3 (H5a) and
with support for military action in Muslim countries at Time 3 (H5b).
Hypothesis 6: e relationship between sources of information about Muslims and support
for public policies harming Muslims will be mediated by perceptions of Muslims as
aggressive and by negative emotions towards Muslims.
Importantly, we controlled for theoretically relevant individual dierences
(American identication, right wing authoritarianism) known to be associated with
negative outgroup attitudes and outgroup harm (Brewer, 2010; Sides & Gross, 2013).
Additionally, we controlled for social desirability as responses on explicit measures
can be aected by such concerns (Hewstone, Rubin, & Willis, 2002).
Study 1: methods
Participants. Participants were 222 undergraduate students (159 female, 58 male, 5
unidentied; Mage =18.76 years) enrolled in introductory communication courses at
a large university in the Midwestern United States, and received course credit for
their participation. Most students self-identied as White (82.9%), followed by Asian
(9.5%), African American (3.6%), Hispanic (1.4%), and other (1.8%). ree students
self-identied as Muslim and four students were suspicious of the study’s hypothe-
ses as assessed by open-ended questions at the end of the study. ese students were
excluded from nal data analysis. Data were collected across two semesters1in three
waves using online surveys at a 3-week interval, over a 9-week period. We will refer
to these three survey assessments as T1, T2, and T3, respectively. ere was a mod-
erate level (15–16%) of attrition on major outcome variables from T1 to T3 across
9 weeks.2Attrition was calculated as the number of participants at T3 compared to
T1.enalresponseratefortheoriginalsamplewas83.81%.Missingdatawere
appropriately handled with multiple imputation methods.
Measures. Table 1 displays the descriptive statistics, alphas, and correlations of key
variables.
Reliance on media-based contact (T1). Participants rated the extent to which
four media sources (newspapers, TV, movies, and Internet) inuenced their opinions
about Muslims on a 1 (not at all)to7(alot) scale (Saleem & Anderson, 2013). ese
items were averaged together.
Journal of Communication (2016) © 2016 International Communication Association 7

Reliance on Direct and Mediated Contact M. Saleem et al.
Table 1 Study 1: Descriptive Statistics and Bivariate Correlations of Hypothesized Variables
(Ns=178–210)
Measure Mean SD Alpha 1 2 3 4 5 6
1. Reliance on media-based
contact (T1)
3.58 1.74 .91 —
2. Reliance on direct contact
(T1)
−.01 .79 .71 .04 —
3. Negative emotions
toward Muslims (T2)
1.74 .91 .97 .32*** −.14* —
4. Perceptions of Muslims as
aggressive (T2)
2.34 .91 .87 .37** −.20* .65*** —
5. Support for civil
restrictions for Muslim
Americans (T3)
2.02 .93 .94 .30*** −.24** .59*** .61* —
6. Support for military
action in Muslim
countries (T3)
2.84 .64 .83 .22* −.16* .48*** .44** .57*** —
Note: Reliance on direct contact is standardized, T1 =Time 1, T2=Time 2, T3=Time 3, SD =standard deviation.
*p<.05. **p<.01.***p<.001.
Reliance on direct contact (T1). Participants were asked (a) the extent to which
they relied on direct personal contact for information about Muslims on a scale rang-
ingfrom1(not at all)to7(alot)(M=3.87; SD =2.28) (Saleem & Anderson, 2013),
(b) the number of Muslims they know (Turner et al., 2007) on a scale ranging from 0
(none)to11(many)(M=4.56; SD =3.53), and (c) of the Muslims they know, how
many would they consider friends (M=3.38; SD =3.26) using the same response
scale.3A single measure of reliance on personal contact was created by standardizing
and averaging the three scores.
Negative emotions toward Muslims (T2). Participants indicated the extent to
which they feel eight emotions toward Muslims (anger, disgust, fury, fear, irritation,
frustration, threat, hostility) on a scale ranging from 1 (not at all)to5(extremely)
(Mackie et al., 2000).
Perceptions of Muslims as aggressive (T2). Participants rated their agreement
with four statements (e.g., “Muslims are violent,” “Most of the terrorists in the world
areMuslim”)onascalerangingfrom1(strongly disagree)to5(strongly agree)(Pratto,
Sidanius, Stallworth, & Malle, 1994).
Support for civil restrictions for Muslim Americans (T3).Participants
responded to six statements (e.g., it is okay for the government to secretly mon-
itor Muslim Americans without their consent or awareness, Muslim Americans
should not be allowed to vote) on a scale ranging from 1 (strongly disagree)to5
(strongly agree) (Saleem, Prot, Anderson, et al., 2015; Saleem, Prot, Cikara, et al.,
2015).
Support for military action in Muslim countries (T3).Participantsindicated
their agreement with seven statements (e.g., the use of drone attacks in Muslim
8Journal of Communication (2016) © 2016 International Communication Association

M. Saleem et al. Reliance on Direct and Mediated Contact
countries is necessary, I do NOT support the use of military action against Muslim
countries-reverse scored) on a scale ranging from 1 (strongly disagree)to5(strongly
agree) (adapted from Henry, Sidanius, Levin, & Pratto, 2005).
Control variables.ControlvariableswereonlyassessedinWave1astraitmea-
sures are unlikely to change in a context of 9 weeks.
IdenticationasanAmerican. Participants responded to four statements (e.g., I feel
proudtobeanAmerican)onascalerangingfrom1(strongly disagree)to7(strongly
agree)(M=5.82; SD =1.13; α=.89; Doosje, Ellemers, & Spears, 1995).
Right wing authoritarianism. Participants responded to 15 statements (e.g., the
“old-fashioned ways” and “old-fashioned values” still show the best way to live) using a
1(Strongly Disagree)to5(Strong Agree)scale(M=2.43; SD =.48; α=.76; Zakrisson,
2005).
Social desirability scale. Participants indicated whether each of the 11 statements
(I’m always willing to admit it when I make a mistake) were true/false in describing
them (Reynolds, 1982). Items indicating social desirability were summed, M=4.51;
SD =2.42; α=.65.
Political orientation. Participants indicated their political orientation on a
scalerangingfrom1(strongly conservative)to7(strongly liberal), M=4.33;
SD =1.69.
Time 1 measures. Time 1 baseline measures of hypothesized mediating variables
(negative emotions toward Muslims and perceptions of Muslims as aggressive)andout-
come variables (support for civil restrictions for Muslim Americans and support for mil-
itary action in Muslim countries) were controlled in nal analyses to examine changes
in these outcomes.
Study1:resultsanddiscussion
Table 1 summarizes the bivariate correlations of key variables used in the model.
Main analysis consisted of testing a path model using Stata 13.1 (StataCorp, College
Station, TX) within the full-information maximum-likelihood (FIML) framework.
Using this approach, we tested all of our hypotheses and the indirect eects of predic-
tor variables passing through two proposed mediators simultaneously in one model.
Overall model t was determined by using the t statistics recommended by Hu and
Bentler (1999). e t indices indicated an excellent t, χ2(22) =36.3253, p=.029,
CFI =0.973, RMSEA =.055 (90% interval from .018 to .086).
Testin g m a i n h y pot h e ses . Figure 1 presents the results of the theoretical SEM model
designed to test H1a–H5b.
Hypotheses 1a and 1b: Reliance on direct contact (T1) inversely predicted changes in negative
emotions towards Muslims (T2) (β=−.22, p<.01), but not changes in perceptions that
Muslims are aggressive (T2) (β=.00, p=.98), providing support for H1b but not H1a.
Hypotheses 2a and 2b: Reliance on media-based contact (T1) positively predicted changes in
negative emotions towards Muslims (T2) (β=.35, p<.001), but not changes in perceptions
that Muslims are aggressive (β=.08, p=.24), supporting H2b but not H2a.
Journal of Communication (2016) © 2016 International Communication Association 9

Reliance on Direct and Mediated Contact M. Saleem et al.
Reliance on direct
contact for info about
Muslims (T1)
Negative emotions
toward Muslims (T2)
Reliance on media for
info about Muslims (T1)
Perceptions that
Muslims are aggressive
(T2)
Support for civil
restrictions for
Muslim Americans
(T3)
Support for military
actions in Muslim
countries (T3)
.19**
.48***
.00 (p = 0.98)
.23**
.20**
.07
.06 (p = 0.36)
(p = 0.29)
Negative emotions
towards Muslims (T1)
.41***
-.13*
Perceptions that Muslims
are aggressive (T1)
Support for civil restrictions
for Muslim Americans (T1)
.37***
.57***
Support for military
actions in Muslim
countries (T1)
Interaction between
reliance on media and
direct contact (T1)
.00
.08 (p = 0.24)
(p = 0.98)
.01 (p = 0.89)
Figure 1 Study 1: e eect of reliance on direct and media-based contact for information
about Muslims (Time 1) on negative emotions toward Muslims and perceptions of Muslims as
aggressive (Time 2) and support for public policies harming Muslims domestically and interna-
tionally (Time 3). Note:Standardizedcoecientsareshownforthehypothesizedpaths.Dotted
lines indicate nonsignicant paths. *p<.05. **p<.01. ***p<.001.
Hypothesis 3a and 3b: Contrary to hypotheses 3a and 3b, the interaction between reliance
on media and direct contact did not signicantly predict changes in perceptions that
Muslims are aggressive (β=.01, p=.89) and changes in negative emotions towards Muslims
(T2) (β=.00, p=.98).
Hypotheses 4a and 4b: Perceptions that Muslims are aggressive (T2) positively predicted
changes in support for civil restrictions for Muslim Americans (T3) (β=.20, p<.01), but not
changes in support for military action in Muslim countries (T3) (β=.06, p=.36), conrming
H4a but not H4b.
Hypotheses 5a and 5b: Negative emotions towards Muslims (T2) positively predicted changes
in support for civil restrictions for Muslim Americans (T3) (β=.23, p<.01) but not changes
in support for military action in Muslim countries (T3) (β=.07, p=.29), conrming H5a but
not H5b. Overall, 56.04% of the variance in support for civil restrictions for Muslim
Americans and 58.09% of the variance in support for military action in Muslim countries was
explained by the predictors included in these models.
Testing mediating eects. InlinewithH6,wetestedthemediatingroleofnegative
emotions toward Muslims and perceptions that Muslims are aggressive between our
predictor variables (reliance on direct and media-based contact) and outcome vari-
ables (support for civil restrictions and military action). We rst isolated the indirect
eect of each mediator in the estimation procedure. Specically, we performed both
the delta and bootstrapping methods to obtain the parameter estimates and standard
10 Journal of Communication (2016) © 2016 International Communication Association

M. Saleem et al. Reliance on Direct and Mediated Contact
errors for the indirect eect of each mediator. e ndings yielded a similar pattern
of results, and thus we report results based on the delta method here.
Reliance on media-based contact (T1) yielded a signicant indirect eect on
changes in support for civil restrictions for Muslim Americans (T3) through negative
emotions (T2) (standardized indirect eect =.04, SE =.0319 p=.025). Similarly,
reliance on direct contact (T1) had a moderately signicant indirect eect on changes
in support for civil restrictions for Muslim Americans (T3) through negative emotions
(T2) (standardized indirect eect =−.03, SE =.017, p=.088). Mediation tests for
support for military action were not performed as the direct link between our
hypothesized mediators (T2) and changes in support for military action in Muslim
countries (T3) was nonsignicant in earlier analyses (H4b and H5b).
Overall, results revealed that reliance on media for information about Muslims
was positively associated with changes in negative emotions toward Muslims but not
signicantly associated with changes in perceptions of Muslims as aggressive. Reliance
on direct contact for information about Muslims was inversely associated with
changes in negative emotions toward Muslims but not signicantly associated with
changes in perceptions of Muslims as aggressive. In other words, reliance on vicarious
and direct contact for information about Muslims might have a stronger inuence in
changing emotions toward, rather than perceptions of, Muslims. Although percep-
tions of Muslims as aggressive were positively associated with support for harming
Muslims domestically and internationally, the latter eect was nonsignicant. Sim-
ilarly, negative emotions toward Muslims were positively associated with changes in
support for harming Muslims domestically and internationally, however, the latter
eect was nonsignicant. ese results suggest that our hypothesized mediators may
predict changes in support for public policies relevant to Muslim Americans better
than public policies relevant to Muslims abroad. e relationship between our predic-
tor variables (reliance on direct and media-based contact) and changes in support for
civil restrictions was mediated by negative emotions, but not by perceptions of Mus-
lims as aggressive. Finally, these results were obtained while controlling for important
theoreticallyrelevantindividualdierencesaswellasTime1baselinemeasures.
Study 2
Study 2 was designed to replicate the ndings of Study 1 using a dierent sample.
Specically, we used a cross-sectional correlational design to test the hypotheses of
Study 1 using a more representative sample of American adults.
Study 2: method
Participants. Data collection occurred through the use of Research Now, a web-based
survey research company, whose panels ensured a more representative sample
than Study 1. Research Now recruits its panel participants through two methods:
self-registration through a web site and proactive recruitment by representatives
of the company through third-party lists. Participants received cash-equivalent
points that can be exchanged for airline miles, gi cards, magazine subscriptions,
Journal of Communication (2016) © 2016 International Communication Association 11

Reliance on Direct and Mediated Contact M. Saleem et al.
Table 2 Study 2: Descriptive Statistics and Bivariate Correlations of Hypothesized Variables
(N=351)
Measure Mean SD Alpha 1 2 3 4 5 6
1. Reliance on media-based
contact
3.40 1.73 .86 —
2. Reliance on direct contact −.004 .83 .79 .23* —
3. Negative emotions toward
Muslims
2.69 1.27 .97 .48** .10 —
4. Perceptions of Muslims as
aggressive
3.12 1.11 .92 .29** −.11* .72** —
5. Support for civil
restrictions for Muslim
Americans
2.79 1.15 .92 .24** −.02 .65** .70** —
6. Support for military
action in Muslim
countries
3.34 .89 .82 .22** −.11* .52** .60**.46** —
Note:SD =standard deviation.
*p<.05. **p<.0.
and other rewards. Quotas were set to achieve equal representation for party iden-
tication (Republicans or Democrats). e nal dataset included 351 participants
(Mage =50.8 years; 54.7% female). Most respondents were White (84.9%), followed
by Asian (6%), Black African American (5.4%), Hispanic (2%), Native American
(1.1%), and Other (0.6%).
Measures. We used all the same measures from Study 1. Table 2 displays descriptive
statistics, including alphas, for all measures used in Study 2.
Study 2: results and discussion
Testin g m a i n h y pot h e ses . Figure 2 presents the results of SEM analysis, using FIML esti-
mations. e model demonstrated excellent t for the data, χ2(11) =32.60, p=.001,
CFI =.979, RMSEA =.075 (90% interval from .046 to .105).
Hypotheses 1a and 1b: Reliance on direct contact was negatively related to perceptions that
Muslims are aggressive (β=−.14, p=.001) but was not signicantly related to negative
emotions towards Muslims (β=.02, p=.610), supporting H1a but not H1b.
Hypotheses 2a and 2b: Consistent with hypotheses 2a and 2b, reliance on media-based
contact was positively related to perceptions that Muslims are aggressive (β=.24 p<.001) and
negative emotions towards Muslims (β=.41, p<.001).
Hypothesis 3a and 3b: H3a and H3b predicted an interaction between reliance on direct and
media-based contact and negative perceptions of and emotions towards Muslims. e
interaction term yielded a signicant eect on perceptions that Muslims are aggressive
(β=.13, p=.003) and negative emotions towards Muslims (β=.11, p=.013). We further
probed these interactions using the simple eects test approach to understand the eect of
12 Journal of Communication (2016) © 2016 International Communication Association

M. Saleem et al. Reliance on Direct and Mediated Contact
Reliance on direct
contact for info
about Muslims
Reliance on media
for info about
Muslims
Negative
emotions towards
Muslims
Perceptions that
Muslims are
aggressive
Support for civil
restrictions for
Muslim Americans
Support for military
actions in Muslim
countries
.41***
.02
(p = .61)
.24***
-.14***
.31***
.36***
.29***
.18**
.13**
Interaction between
reliance on media
and direct contact
.11*
Figure 2 Study 2: e eect of reliance on direct and media-based contact as sources of infor-
mation about Muslims on negative emotions toward Muslims and perceptions of Muslims as
aggressive and support for public policies harming Muslims domestically and internationally.
Note: Standardized coecients are shown for the hypothesized paths. Dotted lines indicate
nonsignicant paths. *p<.05. **p<.01. ***p<.001.
reliance on contact at +1and−1SD of reliance on media. With respect to perceptions that
Muslims are aggressive, reliance on contact was signicant at −1, [t(347) =−3.76, p<.01]
SD but nonsignicant at +1, [t(347) =1.09, p=.27] SD of reliance on media for information
about Muslims (Figure 3). Figure 3 suggests that participants who heavily relied on direct
contact for information about Muslims displayed lower stereotypic beliefs only when they
did not highly rely on media for information about Muslims. Indeed, the dierence between
those who highly relied on direct contact versus those who did not was nonsignicant at
high levels of reliance on media for information about Muslims. With respect to negative
emotions towards Muslims, reliance on contact was nonsignicant at −1, [t(347) =−1.06,
p<.01] SD but marginally signicant at +1, [t(347) =2.00, p=.05] SD of reliance on media
for information about Muslims (Figure 4). Figure 4 suggests that the relationship between
reliance on contact and negative emotions was only signicant for those who highly relied
on media but not for those who had lower levels of reliance on media. ese results are
contradictory to those obtained for the perceptions outcome, however; caution should be
taken when interpreting this interaction given the marginal eect.
Hypotheses 4a and 4b: Conrming hypotheses 4a and 4b, perceptions that Muslims are
aggressive was positively related to support for civil restrictions for Muslim Americans
(β=.36, p<.001) and support for military action in Muslim countries (β=.29, p<.001).
Hypotheses 5a and 5b: Consistent with hypotheses 5a and 5b, negative emotions towards
Muslims was positively associated with support for civil restrictions for Muslim Americans
(β=.31, p<.001) and support for military action in Muslim countries (β=.18, p<.01).
Overall, 57.62% of the variance in support for civil restrictions for Muslim Americans and
47.04% of the variance in support for military action in Muslim countries was explained by
the predictors included in these models.
Journal of Communication (2016) © 2016 International Communication Association 13

Reliance on Direct and Mediated Contact M. Saleem et al.
Figure 3 Study 2: Perceptions of Muslims as aggressive as a function of reliance on media and
reliance on contact for information about Muslims.
Figure 4 Study 2: Negative emotions toward Muslims as a function of reliance on media and
reliance on contact for information about Muslims.
Testing mediating eects. As in Study 1, we tested the signicance of mediating eects
using delta method. We expected that our predictor variables (reliance on direct and
media-based contact) would inuence support for harmful policies (civil restric-
tions and military actions) through two distinct routes: perceptions that Muslims are
aggressive and negative emotions toward Muslims. Results revealed that media-based
contact had a signicant indirect eect on support for civil restrictions for Muslim
Americans through negative emotions (standardized indirect eect =.13, SE =.024,
p<.001) and through perceptions that Muslims are aggressive (standardized indirect
eect =.09, SE =.020, p<.001). Furthermore, media-based contact had a signicant
indirect eect on support for military action in Muslim countries through perceptions
that Muslims are aggressive (standardized indirect eect =.07, SE =.019, p<.001)
and through negative emotions (standardized indirect eect =.07, SE =.024, p<.01).
14 Journal of Communication (2016) © 2016 International Communication Association

M. Saleem et al. Reliance on Direct and Mediated Contact
Mediation results also revealed a signicant indirect eect of reliance on direct
contact on support for civil restrictions for Muslim Americans through perceptions
that Muslims are aggressive (standardized indirect eect =−.05, SE =.018, p<.01),
but not signicantly through negative emotions (standardized indirect eect =.01,
SE =.014, p=.612). Perceptions that Muslims are aggressive signicantly mediated the
relation between direct contact and support for military action in Muslim countries
(standardized indirect eect =−.04, SE =.015, p<.01).
Overall discussion
Two studies examined the extent to which reliance on direct and media-based contact
for information about Muslims inuenced Americans’ stereotypic beliefs, negative
emotions, and support for public policies harming Muslims domestically and inter-
nationally. Across both studies, reliance on media and direct contact were associated
with perceptions of Muslims as aggressive and negative emotions toward Muslims
in the expected direction. However, longitudinal results from Study 1 revealed that
reliance on media and direct contact signicantly predict changes in negative emo-
tions but not changes in perceptions. e interaction between reliance on media and
directcontactwasnonsignicantinStudy1butwassignicantinStudy2.Simple
eects analyses revealed that the positive eects of contact in reducing bias toward
Muslims may be suppressed for participants who also heavily rely on media, likely
due to the negative representation of Muslims in media. is interpretation warrants
caution given the inconsistencies and dierences in samples and design across the two
studies.
Results revealed that perceptions of Muslims as aggressive and negative emo-
tions toward Muslims were positively related to support for public policies harming
Muslims internationally and domestically. However, longitudinal results revealed that
changes in support for civil restrictions were predicted by negative emotions but not
perceptions. Similarly, changes in support for military action were not signicantly
predicted by perceptions or emotions, although both hypothesized mediators were
positively associated with support for military action in Muslim countries. Study
1 mediation results revealed that the eect of reliance on media and direct contact
on changes in support for civil restrictions for Muslim Americans was mediated by
negative emotions but not perceptions. ese results are consistent with previous
studies highlighting the importance of emotions relative to beliefs in predicting out-
group harm (Saleem, Prot, Cikara, et al., 2015; Talaska, Fiske, & Chaiken, 2008) and
policies involving outgroups (Brader, Valentino, & Suhay, 2008). Another important
nding is that reliance on media and direct contact inuenced changes in negative
emotionstoward,butnotperceptionsof,Muslimsasaggressive.Itispossiblethat
information obtained through media and direct contact about Muslims is more
eective in changing emotions toward, rather than perceptions of, Muslims. Indeed,
work on contact theory suggests that contact eects are greater for aective compared
to cognitive outcomes (Pettigrew & Tropp, 2006).
Journal of Communication (2016) © 2016 International Communication Association 15

Reliance on Direct and Mediated Contact M. Saleem et al.
is work has several theoretical and methodological strengths. eoretically
speaking,thisisoneofthefewstudiestodateexaminingthesimultaneous eects of
reliance on direct and media-based sources for information on support for public
policies targeting an outgroup. is is important as individuals oen learn about out-
groups from multiple sources (Mutz & Goldman, 2010). By simultaneously assessing
the extent to which individuals rely on direct and media-based contact for informa-
tionaboutMuslims,wewereabletoexaminetheuniqueandinteractiveeectsof
these predictors on the outcomes of interest. Furthermore, we identied mediating
mechanisms that explain how reliance on media and direct contact can eect support
for public policies targeting outgroups. Overall, this research highlights the diculty
in improving attitudes and public policies toward involving marginalized groups.
Even though the changing demographics of the United States allows individuals
opportunities to interact with members of outgroups and establish positive relations,
mainstream media’s tendency to negatively depict outgroups perpetuates negative
attitudes toward and support for harmful policies targeting these groups.
It is important to note that we assessed support for public policies (military action,
civil restrictions) that are explicitly intended to harm Muslims rather than policies that
passively harm an outgroup (e.g., armative action) or are race-neutral.4Support for
the latter policies is implicitly associated with certain groups (e.g., African Ameri-
cans), however, the policies do not explicitly and exclusively target members of one
outgroup. e policies that we examine in this study are explicitly and exclusively
harming Muslims. is is an important distinction because support for intentionally
harmful policies is a clear representative of active harm, which is oen predicted by
negative emotions and threatening perceptions (Brewer, 2001, 2010).
Methodologically, this work addresses the limitations of past studies in several
ways. e majority of previous media psychology research exploring the eects of
media stereotypes and public policies has been correlational in design, which limits
its causal interpretations (e.g., Mastro & Kopacz, 2006; Tan et al., 2000). Study 1 of this
research used a three-wave longitudinal design to provide stronger evidence for (a) the
direction of the hypothesized relationships, (b) predicting dynamic change from one
time point to another, and (c) testing the underlying mediating mechanisms involved
in understanding the eects of direct and media-based contact on support for public
policies harming outgroups. In addition, Study 2 provided evidence for the general-
izability of Study 1 results by using a more representative adult sample. Moreover, by
controlling for theoretically relevant individual dierences (American identication,
right-wing authoritarianism, social desirability, and political orientation) known to be
associated with outgroup harm, we examined the unique eects of reliance on direct
and media-based contact for information about Muslims.
Finally, this study is one of the few studies in the media psychology literature
to examine attitudes and support for public policies targeting Muslims. is is
alarming when considering the following: (a) the overwhelmingly negative repre-
sentation of Muslims in the American media (e.g., Dixon & Williams, 2015; Nacos
& Torres-Reyna, 2007); (b) the fact that most Americans rely on media as their
16 Journal of Communication (2016) © 2016 International Communication Association

M. Saleem et al. Reliance on Direct and Mediated Contact
primary source of information about Muslims (Nisbet et al., 2009); (c) media’s role
in inuencing negative attitudes toward Muslims is more important than other
sources (Kalkan et al., 2009); (d) most Americans do not have direct contact with
Muslims in their daily lives (e.g., Jung, 2012); and (e) media stereotypes are especially
detrimental for infrequently encountered groups (e.g., Armstrong et al., 1992). us,
becausemostAmericanslackdirectcontactwithMuslims,theyrelyonmediasources
that represent Muslims in a negative light, ultimately increasing negative attitudes
and support for harmful public policies targeting Muslims. Results from this study
suggestthatmediamayhaveasignicantroleinshapingAmericans’recentincrease
in support for civil restrictions for Muslim Americans (e.g., e Brookings Institute,
2011).
Important limitations of this research warrant attention. First, as discussed earlier,
despitetheirtheoreticalsimilaritiesinthecontextofintergroupdynamics,exposure
to and reliance on information are conceptually distinct. Indeed, intergroup contact
theory outlines the hypothesized eects of quantity and quality of contact but not
necessarily reliance on those contacts for information (Allport, 1954). By assessing the
extent to which individuals rely on direct contact and media sources for information
about Muslims, we may have indirectly assessed participants’ own biases regarding
theinuenceofthesesourcesontheirownattitudes.Futureresearchshouldmore
clearly dierentiate the eects of exposure to and reliance on dierence sources in
understanding their unique and combined eects on intergroup outcomes.
Second, the predictor variables used in this research do not isolate the eects of
reliance from valence. Specically, the reliance on direct contact measure included one
positive valence (friendships with Muslims), and two nonvalence items. However, the
reliance on media-based contact measure was nonvalenced. It did not explicitly assess
whether participants were exposed to negative images of Muslims in the media. It is
possible that some individuals who rely on media for information about Muslims are
exposed to positive representations of Muslims, though this seems unlikely based on
research documenting the negative representation of Muslims across American media
outlets (see citations above). Nevertheless, future research should more clearly isolate
theeectsofrelianceandvalencewhenexaminingtheeectsofcontactonintergroup
outcomes, especially because exposure to positive representation of Muslims in news
can decrease support for harmful policies targeting Muslims (Saleem, Prot, Anderson,
et al., 2015).
Overall, these studies address several gaps in the literature. Specically, that direct
and media-based contact can have long-term eects on negative attitudes toward
and support for harmful public policies explicitly and exclusively targeting minor-
ity groups. Whereas direct contact reduces negative attitudes and outgroups harm,
media-basedcontactproducestheoppositeeect.esendingsmakeastrongcase
for the need for more diverse and positive representations of Muslims in mainstream
American media. Simultaneously, there is a need for continued opportunities in the
workplace, in educational settings, in neighborhoods, and elsewhere in the commu-
nity for non-Muslims to interact in meaningful ways with Muslims.
Journal of Communication (2016) © 2016 International Communication Association 17

Reliance on Direct and Mediated Contact M. Saleem et al.
Notes
1 ere were no signicant dierences across the two semesters on key outcomes, thus
these data were combined.
2 ere was no signicant dierence between those who dropped out and those who
remained in the study in terms ofage (t(208) =.915, p=.361), gender (t(208) =.463,
p=.644), and race (ts(208) =1.205, p=.230).
3 Study 2 means and standard deviation for (a) reliance on direct contact, M=3.44,
SD =2.35; (b) number of Muslims you know, M=3.16, SD =3.05; and (c) the number of
Muslimsyouarefriendswith,M=2.67, SD =2.75.
4 Policies such as crime and welfare are inherently race-neutral as they aect all society
members but support for them is heavily inuenced by racial attitudes due to their
implicit association with certain groups (Gilens, 1996).
References
Allport, G. W. (1954). e nature of prejudice. Reading, MA: Addison-Wesley.
Alsultany, E. (2012). ArabsandMuslimsinthemedia:Raceandrepresentationaer9/11.
New York: New York University Press.
Armstrong, G. B., Neuendorf, K. A., & Brentar, J. E. (1992). TV entertainment, news, and
racial perceptions of college students. Journal of Communication,42(3), 153–179.
Atwell Seate, A., & Mastro, D. E. (2015). Media’s inuence on immigration attitudes: An
intergroup threat theory approach. Communication Monographs,83(2), 194 – 213.
Ball-Rokeach, S. J., & Deeur, M. L. (1976). A dependency model of mass media eects.
Communication Research,12, 485 – 510.
Bandura, A. (1986). Self-ecacy. In A. Bandura (Ed.), Social foundations of thought and
action (pp. 390– 453). Englewood Clis, NJ: Prentice Hall.
Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual Review of
Psychology,52(1), 1 –26.
Barlow, F. K., Paolini, S., Pedersen, A., Hornsey, M. J., Radke, H. R., Harwood, J., …Sibley,
C. G. (2012). e contact caveat: Negative contact predicts increased prejudice more than
positive contact predicts reduced prejudice. Personality and Social Psychology Bulletin,
38(12), 1629–1643.
Baumeister, R. F., Bratslavsky, E., Finkenauer, C., & Vohs, K. D. (2001). Bad is stronger than
good. Review of General Psychology,5, 323– 370.
Behm-Morawitz, E., & Ortiz, M. (2012). Race, ethnicity, and the media. In K. Dill (Ed.), e
Oxford handbook of media psychology (pp. 252–264). New York, NY: Oxford University
Press.
Bodenhausen, G., Schwarz, N., Bless, H., & Wänke, M. (1995). Eects of atypical exemplars
on racial beliefs: Enlightened racism or generalized appraisals? Journal of Experimental
Social Psychology,31, 48– 63.
Brader, T., Valentino, N. A., & Suhay, E. (2008). What triggers public opposition to
immigration? Anxiety, group cues, and immigration threat. American Journal of Political
Science,52(4), 959– 978.
Brewer, M. B. (2001). Ingroup identication and intergroup conict. In R. D. Ashmore, L.
Jussim, & D. Wilder (Eds.), Social identity, intergroup conict, and conict reduction
(Vol. 3, pp. 17–41). New York, NY: Oxford University Press.
18 Journal of Communication (2016) © 2016 International Communication Association

M. Saleem et al. Reliance on Direct and Mediated Contact
Brewer, M. B. (2010). Intergroup relations. In R. F. Baumeister & E. J. Finkel (Eds.), Advanced
social psychology: e state of the science (pp. 531–567). New York, NY: Oxford University
Press.
Das, E., Bushman, B. J., Bezemer, M. D., Kerkhof, P., & Vermeulen, I. E. (2009). How
terrorism news reports increase prejudice against outgroups: A terror management
account. Journal of Experimental Social Psychology,45(3), 453–459.
DeFleur, M. L., & Ball-Rokeach, S. (1989). eories of Mass Communication (5th ed.). White
Plains, NY: Longman.
Delli Carpini, M. X., & Keeter, S. (1996). What Americans know about politics and why it
matters.NewHaven,CT:YaleUniversityPress.
Dixon, T. L., & Williams, C. L. (2015). e changing misrepresentation of race and crime on
network and cable news. Journal of Communication,65(1), 24–39.
Doosje, B., Ellemers, N., & Spears, R. (1995). Perceived intragroup variability, group status
and identication. Journal of Experimental Social Psychology,31(5), 410– 436.
Fujioka, Y. (1999). Television portrayals and African American stereotypes: Examination of
contact eects when direct contact is lacking. Journalism & Mass Communication
Quarterly,76, 52–75.
Gaertner, S. L., & Dovidio, J. F. (2000). Reducing intergroup bias: e common ingroup identity
model. Philadelphia, PA: Psychology Press.
Gilens, M. (1996). “Race coding” and white opposition to welfare. American Political Science
Review,90(3), 593 – 604.
Harwood, J., Hewstone, M., Amichai-Hamburger, Y., & Tausch, N. (2012). Intergroup
contact: An integration of social psychological and communication perspectives.
Communication Yearbook,36, 55–102.
Henry, P. J., Sidanius, J., Levin, S., & Pratto, F. (2005). Social dominance orientation,
authoritarianism, and support for intergroup violence between the Middle East and
America. Political Psychology,26, 569 – 583.
Hewstone, M., Rubin, M., & Willis, H. (2002). Intergroup bias. Annual Review of Psychology,
53(1), 575–604.
Hu, L. T., & Bentler, P. M. (1999). Cuto criteria for t indexes in covariance structure
analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A
Multidisciplinary Journal,6(1), 1 –55.
Joyce, N., & Harwood, J. (2012). Improving intergroup attitudes through televised vicarious
intergroup contact: Social cognitive processing of ingroup and outgroup information.
Communication Research,41(5), 627 – 643. doi:10.1177/0093650212447944.
Jung, J. H. (2012). Islamophobia? Religion, contact with Muslims, and the respect for Islam.
Review of Religious Research,54(1), 113–126.
Kalkan, K. O., Layman, G. C., & Uslaner, E. M. (2009). Bands of others? Attitudes toward
Muslims in contemporary American society. Journal of Politics,71(3), 847 – 862.
Kteily, N., Bruneau, E., Waytz, A., & Cotterill, S. (2015). “e ascent of man”: A theoretical
and empirical case for blatant dehumanization. Journal of Personality and Social
Psychology,109(5), 901 –931.
Lowery, B. S., Unzueta, M. M., Knowles, E. D., & Go, P. A. (2006). Concern for the in-group
and opposition to armative action. Journal of Personality and Social Psychology,90(6),
961–974.
Journal of Communication (2016) © 2016 International Communication Association 19

Reliance on Direct and Mediated Contact M. Saleem et al.
Mackie, D. M., Devos, T., & Smith, E. R. (2000). Intergroup emotions: Explaining oensive
action tendencies in an intergroup context. Journal of Personality and Social Psychology,
79(4), 602–616.
Mastro, D. (2009). Eects of racial and ethnic stereotyping. Media Eects: Advances in eory
and Research,3, 325 –341.
Mastro, D., Behm-Morawitz, E., & Ortiz, M. (2007). e cultivation of social perceptions of
Latinos: A mental models approach. Media Psychology,9(2), 347 – 365.
Mastro, D. E., & Kopacz, M. A. (2006). Media representations of race, prototypicality, and
policy reasoning: An application of self-categorization theory. Journal of Broadcasting &
Electronic Media,50(2), 305–322.
Mastro, D. E., & Tropp, L. R. (2004). e eects of interracial contact, attitudes, and
stereotypical portrayals on evaluations of Black television sitcom characters.
Communication Research Reports,21(2), 119–129.
Mummendey, A., & Otten, S. (2001). Aversive discrimination. In R. Brown & S. L. Gaertner
(Eds.), Handbook of social psychology: Intergroup processes (pp. 112–132). Oxford,
England: Blackwell.
Mutz, D. C., & Goldman, S. K. (2010). Mass media. In J. F. Dovidio, M. Hewstone, P. Glick, &
V. M. Esses (Eds.), Handbook of prejudice, stereotyping, and discrimination (pp. 241–258).
London, England: Sage.
Nacos, B. L., & Torres-Reyna, O. (2007). Fueling our fears: Stereotyping, media coverage, and
public opinion of Muslim Americans.NewYork,NY:Rowman&Littleeld.
Nisbet, E. C., Ostman, R., & Shanahan, J. (2009). Public opinion toward Muslim Americans:
Civil liberties and the role of religiosity, ideology, and media use. In A. Sinno (Ed.),
Muslims in western politics (pp. 161– 199). Bloomington, IN: Indiana University Press.
Ortiz, M., & Harwood, J. (2007). A social cognitive theory approach to the eects of mediated
intergroup contact on intergroup attitudes. Journal of Broadcasting & Electronic Media,
51, 615–631.
Paolini, S., Harwood, J., & Rubin, M. (2010). Negative intergroup contact makes group
memberships salient: Explaining why intergroup conict endures. Personality and Social
Psychology Bulletin,36, 1723 – 1738.
Paolini, S., Harwood, J., Rubin, M., Husnu, S., Joyce, N., & Hewstone, M. (2014). Positive and
extensive intergroup contact in the past buers against the disproportionate impact of
negative contact in the present. European Journal of Social Psychology,44(6), 548–562.
Pettigrew, T. F., & Tropp, L. R. (2006). A meta-analytic test of intergroup contact theory.
Journal of Personality and Social Psychology,90, 751– 783.
Power, J. G., Murphy, S. T., & Coover, G. (1996). Priming prejudice: How stereotypes and
counter-stereotypes inuence attribution of responsibility and credibility among
ingroups and outgroups. Human Communication Research,23, 36 –58.
Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). Social dominance orientation:
A personality variable predicting social and political attitudes. Journal of Personality and
Social Psychology,67(4), 741– 763.
Ramasubramanian, S. (2011). e impact of stereotypical versus counterstereotypical media
exemplars on racial attitudes, causal attributions, and support for armative action.
Communication Research,38, 497 – 516. doi:10.1177/0093650210384854.
Ramasubramanian, S. (2013). Intergroup contact, media exposure, and racial attitudes.
Journal of Intercultural Communication Research,42, 54–72.
20 Journal of Communication (2016) © 2016 International Communication Association

M. Saleem et al. Reliance on Direct and Mediated Contact
Reynolds, W. M. (1982). Development of reliable and valid short forms of the
Marlowe-Crowne Social Desirability Scale. JournalofClinicalPsychology,38, 119 – 125.
Riek, B. M., Mania, E. W., & Gaertner, S. L. (2006). Intergroup threat and outgroup attitudes:
A meta-analytic review. Personality and Social Psychology Review,10(4), 336–353.
Saleem, M., & Anderson, C. A. (2013). Arabs as terrorists: Eects of stereotypes within
violent contexts on attitudes, perceptions, and aect. Psychology of Violence,3(1), 84–99.
Saleem, M., Prot, S., Anderson, C. A., & Lemieux, A. F. (2015). Exposure to Muslims in media
and support for public policies harming Muslims. Communication Research.
doi:10.1177/0093650215619214.
Saleem, M., Prot, S., Cikara, M., Lam, B. C. P., Anderson, C. A., & Jelic, M. (2015). Cutting
Gordian knots: Reducing prejudice through attachment security. Personality and Social
Psychology Bulletin,41(11), 1560 – 1574. doi:10.1177/0146167215601829.
Schiappa, E., Gregg, P. B., & Hewes, D. E. (2005). e parasocial contact hypothesis.
Communication Monographs,72, 92 –115.
Shaheen, J. (2009). Reel bad Arabs: How Hollywood vilies a people. Northampton, MA: Olive
Branch Press.
Sides, J., & Gross, K. (2013). Stereotypes of Muslims and support for the war on terror.
Journal of Politics,75, 583– 598.
Stephan, W., & Stephan, C. W. (2000). An integrated threat theory of prejudice. In S. Oskamp
(Ed.), Reducing prejudice and discrimination (pp. 23–45). Hillsdale, NJ: Erlbaum.
Talaska, C. A., Fiske, S. T., & Chaiken, S. (2008). Legitimating racial discrimination:
Emotions, not beliefs, best predict discrimination in a meta-analysis. Social Justice
Research,21(3), 263– 296.
Tan, A. S., Fujioka, Y., & Tan, G. (2000). Television use, stereotypes of African Americans and
opinions on armative action: An aective model of policy reasoning. Communication
Monographs,67, 362 –371.
e Brookings Institute. (2011). What it means to be American.Retrievedfrom
http://www.brookings.edu/˜/media/research/les/reports/2011/9/06-american-attitudes/
0906_american_attitudes.pdf.
Tukachinsky, R., Mastro, D., & Yarchi, M. (2015). Documenting portrayals of race/ethnicity
on primetime television over a 20-year span and their association with national-level
racial/ethnic attitudes. JournalofSocialIssues,71(1), 17–38.
Turner, R. N., Hewstone, M., & Voci, A. (2007). Reducing explicit and implicit prejudice via
direct and extended contact: e mediating role of self-disclosure and intergroup anxiety.
Journal of Personality and Social Psychology,93, 369– 38.
Zakrisson, I. (2005). Construction of a short version of the Right-Wing Authoritarianism
(RWA) Scale. Personality and Individual Dierences,39(5), 863–872.
Journal of Communication (2016) © 2016 International Communication Association 21