The Problem With Protests: Emotional Effects of Race-Related News Media

Abstract

It is well documented that news media’s coverage of social unrest is sensationalized; however, our knowledge is limited in understanding how the intersection of race with depictions of social unrest influences emotional responses to this content. By applying assumptions from the protest paradigm and intergroup emotions theory, the current set of studies experimentally examines this relationship. Results indicate that racialized news images of dramatized social unrest provoke heightened, complex group-based affective responses that vary based on aspects of psychological group identification among audiences. These outcomes suggest that journalistic practices, whether or not intentionally, may exacerbate race relations regarding social change.

Keywords

collective action intergroup emotion protest paradigm race

Social unrest, whether expressed in the form of nonviolent opposition or aggressive resistance, is an ever-present reality in society and an elemental aspect of democratic cultures. Of course, an inescapable feature of such civic action is group-level conflict, such that the rights of disaffected and/or underserved groups are seen, whether or not accurately, to threaten the standing and opportunities of dominant groups in society and of the status quo, more generally. Although this tension may propel social change, it can also lead to the vilification of already marginalized communities (e.g., racial or ethnic groups). Notably, the manner in which news coverage depicts social unrest is likely to contribute to this outcome. As a result of its tendency to delegitimize groups involved in social unrest (Boyle et al., 2012), news coverage has the potential to not only influence views on social issues but also affect the group-level emotions, or emotions derived from audiences, potentially exacerbating group-level antipathy. Although this is consequential for any group engaged in collective action, it is particularly problematic for at-risk and underserved populations. The present set of experimental studies investigates this relationship, applying insights from the protest paradigm (e.g., McLeod, 2007) and intergroup emotions theory (IET; for example, Mackie & Smith, 2015). More specifically, the influence of exposure on news coverage intersecting depictions of race (Black, White) and social unrest (protest, riot), on the emotional responses of audiences is examined.

Protest Paradigm

Research examining news coverage of protests has documented a trend in the characterization of social unrest such that dramatized coverage (e.g., violence, group tactics) is privileged over explication of the issues or articulation of the group’s goals (Boyle et al., 2012; McLeod, 2007). Known as the protest paradigm (McLeod, 2007), this pattern in news coverage serves the functional purpose of garnering interest among audiences and reinforcing dominant perspectives, while simultaneously disadvantaging groups’ attempts at collective action. Thus, rather than providing a platform for the enhancement and legitimization of diverse views, news coverage of protest is likely to diminish and harm such efforts (Weaver & Scacco, 2013). As McLeod (2007, p. 185) asserts, the messages and images used in news coverage of protests delegitimizes and “disparages protesters and hinders their role as vital actors on the political stage.”

By using “routinized templates for creating protest stories” coverage is not only constrained, but the status quo is reinforced; ultimately serving dominant perspectives (McLeod, 2007, p. 186). As Boyle et al. (2005, pp. 638–639) find in their analysis of nearly 40 years of news coverage of protest activity in the United States, “the more protest groups threaten the status quo, the more harshly they will be treated by the media.” Indeed, research documenting the content patterns identified in the protest paradigm suggest that, whether or not intentionally or consciously, news outlets act to exert a social control function that both inhibits social change and discredits protesters (Boyle et al., 2005, 2012). This occurs through the practice of relying on “official” sources over protesters’ issues, focusing on the appearance of protesters rather than the circumstance that caused the unrest, de-emphasizing the situational context surrounding the issues, focusing on conflict and/or disorder, and downplaying the effectiveness of the social actors (Brasted, 2005a, 2005b; Weaver & Scacco, 2013). Furthermore, research finds that issues related to social unrest that are seen as extreme or deviant are treated even more critically and severely in the media (Boyle et al., 2005; Weaver & Scacco, 2013).

These characterizations can have a powerful influence on audience members’ views of these issues and the protesters themselves (Boyle et al., 2012; Weaver & Scacco, 2013). Moreover, the visual images commonly accompanying these stories can exacerbate these responses (Arpan et al., 2006; Powell et al., 2015). Indeed, research finds that the pictures presented with news reports serve as meaningful sources of information for consumers which, when presented in a manner consistent with the protest paradigm (i.e., emphasizing disorder and violence) can further damage and delegitimize the social actors and their issues, particularly when the issue is of interest to the audience member (Arpan et al., 2006). Moreover, visual images appear to play a dominant role (over text) in determining behavioral responses, with anger and fear mediating this relationship (Powell et al., 2015).

Ultimately then, the narrative construction common within coverage of social unrest, problematizes the protesters and trivializes their positions through the characterization of these individuals as deviant and threatening. In characterizing social unrest in this way, media coverage may create a muddled narrative such that protests are seen as violent or even as indistinguishable from a riot. The two separate forms of social unrest deserve clear and concise conceptualization, as there are specific and distinctive features which differentiate protests from riots (Betti, 2016). A protest is an organized public demonstration in which strong objection is voiced regarding policy, practice, or actions of institutions, individuals, or other entities. A protest can take many forms including but not limited to a sit-in, march, rally, boycott, or hunger strike; none of which involve violent actions among participants. If violent acts occur during a protest, this action shifts the event to that of a riot. Riots are violent disruptions by a crowd that may result in damage to or destruction of property, potential looting, and harm to individuals, ranging from arrests to death.

The tendency of news media to conflate different forms of collective action events can serve to delegitimize social actors and obscure their issues. This is likely to be uniquely damaging to already marginalized or at-risk communities such as racial and/or ethnic groups. In light of the collective movements currently at the forefront of the social stage ranging from Black Lives Matter, to the Immigration Rights and Sanctuary City Movement, to the Indigenous Environmental Network, this is particularly problematic. If the tendency in news depictions of social protest is to emphasize conflict and violence over social criticism, it is likely that alongside delegitimizing the subject matter, this coverage may exert considerable influence over the emotions associated with the issue and its actors. Here, IET offers important insights.

IET

Grounded in social identity and self-categorization-based theorizing, IET considers the unique nature of emotions when experienced at the group level as opposed to the individual level (Mackie & Smith, 2015). Differentiating group-level from individual-level emotions is meaningful given that research demonstrates the potential for group-based evaluations to evoke distinct interpretations of the social world (and the people in it), which then may elicit specific emotions associated with unique and specific behavioral responses (see Caprariello et al., 2009).

The activation and operation of group-level emotions are consistent with processes articulated in social identity frameworks, such that group-level emotions can be evoked based on subtle or blatant category activation as well as based on appraisal of group-relevant contexts and phenomenon (Mackie & Smith, 2015). The distinct emotional experience that emerges from such group-based categorization varies based on the salient group identity. In other words, group-based emotional profiles differ based on the relevant social identity such that “merely activating different social categorizations result[s] in different emotional experiences” for the same individual (Mackie & Smith, 2015, p. 265). Furthermore, these group-based emotions are distinct in form and function from individual-level emotions.

Although emotions readily occur at the group level (e.g., Mackie & Smith, 2015; Smith et al., 2007), until fairly recently, scholarship had focused on emotions as distinctly individual-level experiences. Theory and research in the domain of IET suggest that we possess and experience group-specific emotions in much the same way as we might hold cognitions about different groups (e.g., Mackie et al., 2000). These group-level emotions serve as a motivating force that can meaningfully influence behaviors toward different groups in society, including both outgroups and ingroups. Indeed, empirical evidence supports this assertion, demonstrating the impact of intergroup emotions (vs. individual emotions) on prejudicial attitudes and behaviors as well as on tendencies to attack or avoid outgroups (Leonard et al., 2011; Maitner et al., 2006, 2007; Miller et al., 2004; Ray et al., 2008).

Group-level emotions may also prompt constructive outcomes when favorable contact experiences or socialization lead to positive emotions toward a group. Such favorable group-level emotions would be expected to meaningfully improve intergroup relationships by promoting specific intergroup behaviors in-line with the emotional association (Mackie et al., 2008). Ultimately, this research indicates that different group-level emotions prompt differentiated and predictable actions which are both consistent with the emotional norm and unique from individual-level emotions both qualitatively as well as in terms of the intensity of the emotion (Mackie & Smith, 2015; Moons et al., 2009). As Mackie and Smith (2015, p. 268) articulate, “What makes it clear that such emotions are a group-level phenomenon . . . is that they are based not on appraisals of consequences for the self but rather on appraisals made in relation to the ingroup.”

Furthermore, research on intergroup emotions indicates that distinct behaviors are associated with specific group-level emotions (see Maitner et al., 2006). For example, intergroup anger provokes confrontational actions, including attacking and directly harming the target group (Yzerbyt et al., 2003). Alternatively, group-level fear prompts avoiding and excluding behaviors, rather than direct confrontation and attack (Mackie et al., 2009). However, intergroup guilt encourages compensatory or reparative behaviors toward the wronged or mistreated outgroup (Doosje et al., 1998). Different still, intergroup satisfaction signals the maintenance of current behaviors toward the outgroup (Maitner et al., 2006). Thus, specific group-level emotions are tied to unique and particular group-based actions. To further illustrate, group-level fear has been found to prompt support for restrictive public policy (i.e., avoidance) against the out-group as opposed to attacking behaviors toward the group (e.g., Skitka et al., 2004). Similarly, group-based anger has been found to generate confrontational (i.e., attack and aggress) reactions (Iyer et al., 2007; Smith et al., 2007) but not avoidance or reparations (Mackie et al., 2000).

Group Membership and Intergroup Emotions

Certainly then, group identity is likely to be central to the development and expression of intergroup emotions and related actions (Mackie et al., 2004; Yzerbyt et al., 2003). Research has found that compared with those holding low levels of group identification, individuals high in group identification experience more pronounced and intense intergroup emotions (Mackie et al., 2004; Yzerbyt et al., 2003), which then encourage the production of the emotion-specific action tendency (Yzerbyt et al., 2003).

Although at first blush a number of group memberships (including political, ideological, racial, etc.) may seem central, research indicates that social dominance orientation (SDO; Pratto et al., 1994) can be a stronger predictor of both group-related social policy preferences (Pratto et al., 1994) and attitudes toward underserved groups including the following: undocumented immigrants (e.g., Bassett, 2010; Danso et al., 2007); women (e.g., Heaven, 1999); lesbian, gay, bisexual, and transgender (LGBT) communities (e.g., Whitley & Ægisdόttir, 2000); and other marginalized groups (see Ho et al., 2012). In the context of news coverage of racialized social unrest, then, group membership might be best understood in terms of SDO; in particular, the subdimensions of group dominance and anti-egalitarianism (Ho et al., 2015).

SDO is inherently a group phenomenon which “entails individuals’ desire for their social groups to dominate and subordinate other groups” (Danso et al., 2007, p. 1114). Although not a traditional group membership in the context of IET research, SDO reflects group-based identities tied to beliefs about group hierarchy and opposition to group equality (Ho et al., 2012). The two complementary but distinct psychological orientations within SDO predict different intergroup outcomes: with group dominance related to “aggressive intergroup behavior,” “zero-sum competition,” and “old-fashioned prejudice” and anti-egalitarianism predicting “subtle forms of intergroup biases,” “perpetuating systems of group-based inequalities,” and “modern prejudice” (Ho et al., 2012, pp. 594–595). As such, it is reasonable to conclude that adherence to these psychological categories may better predict intergroup emotions in the context of news coverage intersecting depictions of race with social unrest, then other group identities.

News Coverage and Intergroup Emotions

Certainly then, the types of media characterizations associated with news coverage of social unrest have the potential not only to delegitimize the group, as research on the protest paradigm suggests (e.g., Boyle et al., 2012; McLeod, 2007), but also to promote distinctly disadvantageous feelings toward the groups involved; ultimately encouraging harmful intergroup outcomes. Although limited, the small number of studies testing the role of media in intergroup emotions-related processes and effects supports this claim. This research indicates that exposure to intergroup threat in the media can encourage intergroup contempt (which is associated with avoidance and exclusion behaviors), prompting restrictive policy preferences, and information sharing in support of such preferences (Atwell Seate & Mastro, 2017). Similarly, TV consumption has been found to be associated with increased group-level anger toward racial/ethnic groups in the United States (Atwell Seate et al., 2018). Finally, although not directly testing IET, Ramasubramanian (2010) found that group-based emotions prompted by television use, mediate the influence of cognitions on group-level attitudes and policy positions.

The Current Set of Studies

When taken together, theoretical assumptions and empirical evidence rooted in the protest paradigm and IET suggest that a uniquely challenging environment may exist when underserved communities, such as racial or ethnic groups, attempt to engage in collective action or social change efforts. In particular, it appears that the coupling of depictions of race with protest paradigm constructions is likely to evoke group-based emotional responses that are distinctly harmful for the efforts of these groups. Research additionally suggests that consumers’ SDO may affect group-level emotional responses in this context, although the specific influence of group dominance orientation versus anti-egalitarian orientation is less clear. From this integrated perspective, the following Hypothesis and Research Question were proposed:

H1 (a–e): News coverage of social unrest (protest vs. riot), race of the depicted participants (Black vs. White) and the social dominance of the consumer will interact in predicting affective responses to the content such that as social dominance orientation increases, exposure to Black riots will result in increased (a) agitation, (b) offense, (c) shame, (d) disgust, and (e) fear.

RQ1 (a–d): When exposed to news depictions of race and social unrest, will social dominance orientation influence feelings of (a) inspiration, (b) pride, (c) hope, and (d) guilt?

Method: Study 1

Participants

Undergraduate students (N = 183) from a large western university took part in this study on a voluntary and anonymous basis. Among these, 33% were White (N = 61), 28% were Asian (N = 51), 17% were Latino/a (N = 31), 16% were of unknown, mixed ethnic background (N = 30), and less than 4% (N = 7) were of another racial/ethnic background (e.g., Middle Eastern). Given the intergroup nature of the current predictions, it was not possible to conduct the distinct, relevant analyses with Black participants, as only three (2%) took part in the study. As such, they were dropped from analyses. Sixty-nine percent were female (N = 127), 30% were male (N = 54), and 1% did not report gender identity (N = 2). The average age of the sample was 19.85 (SD = 1.61).

Procedure

Participants were told that they were taking part in a study examining perceptions of current events in news articles. They were informed that their responses were voluntary and anonymous and that they could quit the study at any time, without penalty. The experimental testing session was conducted in a controlled lab setting, using hard copy questionnaires and hard copies of news articles. Data collection took place between February and April of 2018. Participants were randomly assigned to one of the four experimental conditions varying the type of social unrest (protest or riot) and race (Black or White) of the individuals pictured in the article. Trained undergraduate research assistants managed aspects of the experiment including handling hard copy materials and debriefing participants after the conclusion of the study. Completion of the study by participants ranged from 20 to 35 min.

Experimental Manipulation

Consistent with existing research, the news articles used for the experimental manipulation were adapted from actual news articles appearing in U.S. news outlets (e.g., Dixon, 2006). Within both of the social unrest conditions (riot or protest), news articles were accompanied by photographs depicting either all Black or all White individuals; with the photo being the only racial cue within the entire news story. Within each photograph that accompanied each news story all actors were of the same race, either all White or Black individuals. In the protest condition, the image featured individuals marching outside in a large group with nonspecific protest signs (e.g., signs read “Fair Treatment 4 All!”). The text of the news articles in the protest condition was exactly the same in both race conditions. In the article, the term protesters was specifically used in reference to the individuals involved and the topic under protest was intentionally indistinct and revolved around the issue of economic and structural inequalities in society. In the riot condition, images featured an action photo including large groups with individuals destroying property and looting. Each accompanying image depicted all Black or all White actors and each image depicted multiple actors within the single photo, all of the same race. In the text of the news article, the individuals were specifically referred to as rioters and consistent with the protest condition, the article addressed economic conditions and structural inequalities in society. The mention of economic and structural inequalities was used to emphasize the theme of collective action and was intentionally designed so that no specific or real-world example of social unrest was evoked.

The text of the articles in the riot condition was the same in both race conditions. There was no explicit mention of race or specific social groups in the text of the news articles for any condition. In addition, there were no mention of current or past protests (e.g., March for Our Lives or Black Lives Matter) within the news conditions used for this study. However, during the time of data collection, multiple protests and riots were taking place in the United States including March for Our Lives, The Women’s March, various immigration protests, and several unnamed protests targeting the current U.S. administration. In sum, the experimental conditions varied in terms of levels of social unrest (riot or protest) and the image of the racial group involved in the social unrest (Black or White).

Pilot Testing

To ensure that the attributes of interest were appropriately reflected in the news articles a pilot test was conducted on a separate sample of undergraduate communication students (N = 72) from the same public institution. Participants, individually and in a laboratory setting, read a randomly assigned news condition and answered a hard copy questionnaire. The majority of the participants correctly identified the race of the individuals depicted in the news articles (88%, N = 64) and the vast majority of participants also correctly identified the specific action (riot or protest) taking place (96%, N = 69). On a 5-point scale from 1 “not at all” to 5 “very” participants also rated if the conditions presented were distinguishable between riots and protests on specific variables including how peaceful, civil, organized, scary, and intimidating each news article was. Analysis of variance (ANOVAs) was used to compare the four conditions. As expected, the Black protest (M = 3.47, SD = 1.26) and White protest (M = 3.25, SD = 1.34) were rated as significantly more peaceful than the Black riot (M = 1.68, SD = .67) and White riot (M = 1.06, SD = .24) conditions, F(3, 71) = 26.43, p < .01. Likewise, the Black protest (M = 4.11, SD = .88) and White protest (M = 3.56, SD = 1.15) were rated as significantly more civil than the Black riot (M = 1.53, SD = .70) and White riot (M = 1.72, SD = 1.02) conditions, F(3, 71) = 35.01, p < .01. The Black protest (M = 3.47, SD = 1.07) and White protest (M = 3.25, SD = 1.06) were also rated as significantly more organized than the Black riot (M = 1.63, SD = .90) and White riot (M = 1.50, SD = .79) conditions, F(3, 71) = 21.39, p < .01. Piloting testing also revealed that the Black riot (M = 3.47, SD = .96) and White riot (M = 3.83, SD = .96) were rated as significantly scarier than the Black protest (M = 1.84, SD = .90) and White protest (M = 2.00, SD = 1.15) conditions, F(3, 71) = 18.57, p < .01. Finally, the Black riot (M = 3.68, SD = 1.11) and White riot (M = 3.72, SD = 1.13) were rated as significantly more intimidating than the Black protest (M = 2.42, SD = 1.07) and White protest (M = 2.56, SD = 1.15) conditions, F(3, 71) = 7.21, p < .01.

To ensure that news stories were consistent along other attributes that may influence readers’ perceptions and responses, participants rated the articles on a 5-point scale from 1 “not at all” to 5 “very,” in terms of, typicality (i.e., representing similar characteristics of a “real” news article), ease to follow, as well as the degree to which articles were perceived to be interesting and engaging. ANOVA revealed no significant differences in these attributes based on condition: typicality of the news article, F(3, 71) = 1.27, p = .291; ease to follow, F(3, 71) = 1.02, p = .389; interest in the news article, F(3, 71) = .678, p = .568; and the engagingness of the news article, F(3, 71) = 1.01, p = .393.

Moderating Variables

Social dominance

SDO measures (Ho et al., 2015; Pratto et al., 1994) were used to create two known, psychologically distinct subscales (see Ho et al., 2015) assessing the following: (a) group dominance (α = .84, M = 2.39, SD = 1.42) and (b) anti-egalitarianism (α = .86, M = 2.21, SD = 1.08). The four-item measure of group dominance included statements such as, “Some people are just more worthy than others” and “Some people are just inferior to others.” The six-item measure of anti-egalitarianism included statements such as, “If people were treated more equally, we would have fewer problems in this country” and “We should try to treat one another as equals as much as possible.” Response options ranged from “strongly agree” to “strongly disagree” on a 7-point scale, with higher numbers reflecting greater adherence to SDO.

Dependent Variables

Group emotions

Consistent with Mackie et al. (2000), group-level emotions were assessed with the following single item measures: agitated, offended, ashamed, disgusted, afraid, inspired, proud, hopeful, guilt. Filler items also were included to help conceal the intent of the questions. Participants were asked to rate the extent to which they felt each of the emotions after reading the article, from “not at all” (1) to “very strongly” (7).

Results: Study 1

H1: Social Unrest in the News, Social Dominance, and Negative Affect

H1a–e predicted a three-way interaction between news depiction of social unrest (protest vs. riot), race of the depicted participants (Black vs. White), and the social dominance of the consumer in predicting (a) agitation, (b) offense, (c) shame, (d) disgust, and (e) fear. Both the group dominance and anti-egalitarian subscales of SDO were examined, as they would be expected to uniquely influence outcomes, in possibly distinct ways.

Univariate general linear model (GLM) revealed a significant three-way interaction between news condition, race of the depicted participants, and group dominance in predicting both agitation, F(3, 175) = 3.69; p < .025; partial $η^{2} = . 06$ , and offense, F(3, 175) = 3.73; p < .025; partial $η^{2} = . 06$ . As Figure 1 reveals, as group dominance increases agitation sharply increases for those exposed to the Black riot news story. Furthermore, as group dominance increases, agitation decreases when exposed to the White riot, revealing a crossover interaction in the riot condition. In terms of feeling offended, Figure 2 demonstrates that, again, as group dominance increases so too do feelings of being offended when exposed to the Black riot news article. Here, social dominance had no influence on feelings of offense in any other news condition.

Figure 1.

Three-way interaction between news condition (riot vs. protest), race condition (Black vs. White) and group dominance in predicting agitation.

Figure 2.

Three-way interaction between news condition (riot vs. protest), race condition (Black vs. White) and group dominance in predicting offense.

Univariate GLM revealed a significant three-way interaction between news condition, race of the depicted participant, and the anti-egalitarianism subscale in predicting shame, F(3, 174) = 4.93; p < .01; partial $η^{2} = . 08$ , and disgust, F(3, 174) = 2.71; p < .05; partial $η^{2} = . 05$ . As illustrated in Figure 3, shame responses were most pronounced in the riot condition with increasing anti-egalitarianism associated with increasing feelings of shame when exposed to the White riot condition. Similarly, as Figure 4 demonstrates, disgust was more strongly associated with exposure to the riot articles; and as anti-egalitarianism increased, so too did feelings of disgust when exposed to the White riot.

Figure 3.

Three-way interaction between news condition (riot vs. protest), race condition (Black vs. White) and anti-egalitarianism in predicting shame.

Figure 4.

Three-way interaction between news condition (riot vs. protest), race condition (Black vs. White) and anti-egalitarianism in predicting disgust.

No other significant three-way interactions emerged in tests of H1. However, a main effect of news condition emerged, F(3, 179) = 4.43, p < .01 in fear responses. Specifically, exposure to the riot articles (irrespective of depictions of Black [M = 3.04, SD = 1.73] or White [M = 3.07, SD = 1.70] rioters) prompted significantly more fear than exposure to the protest articles (irrespective of depictions of Black [M = 2.22, SD = 1.33] or White [M = 2.23, SD = 1.40] protesters). See Table 1 for descriptive statistics for H1a–e.

Table 1.

Descriptive Statistics for Hypothesis 1a–e.

	N	M	SD	SE	95% CI
Affective Responses	N	M	SD	SE	Lower	Upper
Agitated
Black protest	46	1.89	1.354	.200	1.49	2.29
White protest	48	2.02	1.537	.222	1.57	2.47
Black riot	46	2.65	1.649	.243	2.16	3.14
White riot	43	3.30	1.780	.271	2.75	3.85
Total	183	2.45	1.666	.123	2.21	2.69
Offended
Black protest	46	1.30	.726	.107	1.09	1.52
White protest	48	1.29	.849	.123	1.05	1.54
Black riot	46	1.87	1.327	.196	1.48	2.26
White riot	43	1.79	1.301	.198	1.39	2.19
Total	183	1.56	1.102	.081	1.40	1.72
Ashamed
Black protest	46	1.70	1.331	.196	1.30	2.09
White protest	48	1.69	1.307	.189	1.31	2.07
Black riot	46	2.67	1.874	.276	2.12	3.23
White riot	43	2.26	1.733	.264	1.72	2.79
Total	183	2.07	1.617	.120	1.84	2.31
Disgusted
Black protest	46	1.80	1.408	.208	1.39	2.22
White protest	48	1.79	1.368	.197	1.39	2.19
Black riot	46	2.65	1.741	.257	2.14	3.17
White riot	43	2.58	1.735	.265	2.05	3.12
Total	183	2.20	1.609	.119	1.96	2.43

Note. higher mean score indicates stronger emotion. CI = confidence interval.

Research Question: Social Unrest in the News and Positive Affect

The Research Question probed the influence of SDO on feelings of (a) inspiration, (b) pride, (c) hope, and (d) guilt, when exposed to news depictions of race and social unrest. A three-way interaction did not emerge in GLM analyses of news condition, race of depicted participants, and social dominance for any of the positive feelings. However, there was a main effect of news condition on inspiration, F(3, 179) = 14.85, p < .01; pride, F(3, 179) = 19.52, p < .01; and hope, F(3, 179) = 17.27, p < .01. Regardless of the race of the depicted participants, news consumers were more inspired, prouder, and more hopeful when exposed to news coverage of protests than of riots. See Table 2 for descriptive statistics.

Table 2.

Descriptive Statistics for Positive Emotions by News Condition.

	N	M	SD	SE	95% CI
	N	M	SD	SE	Lower	Upper
Inspired
Black protest	46	3.98	2.124	.313	3.35	4.61
White protest	48	4.00	1.924	.278	3.44	4.54
Black riot	46	2.17	1.568	.231	1.71	2.64
White riot	43	2.26	1.428	.226	1.80	2.71
Total	183	3.13	1.995	.147	2.83	3.42
Proud
Black protest	46	3.50	2.106	.310	2.87	4.13
White protest	48	3.98	1.862	.269	3.44	4.52
Black riot	46	2.07	1.511	.223	1.62	2.51
White riot	43	1.67	1.085	.165	1.34	2.01
Total	183	2.84	1.937	.143	2.55	3.12
Hopeful
Black protest	46	4.28	1.905	.281	3.72	4.85
White protest	48	4.58	1.966	.284	4.01	5.15
Black riot	46	2.72	2.029	.299	2.11	3.32
White riot	43	2.23	1.525	.233	1.76	2.70
Total	183	3.49	2.109	.156	3.18	3.79

Note. higher mean score indicates stronger emotion. CI = confidence interval.

Brief Study 2, Rationale

Results from Study 1 offer preliminary support for predictions stemming from IET and the protest paradigm, indicating that racialized news images of dramatized social unrest provoke heightened, complex emotional responses that vary based on aspects of psychological group identification among audiences. However, this relationship was tested with a sample of college students. Although the current study’s theoretically derived predictions, which focused on psychological factors (i.e., group dominance and anti-egalitarianism) as opposed to the typical group identities often used in these contexts (e.g., racial identity, political ideology), should not be meaningfully constrained based on this sample, it undermines generalizability, given the lack of varied demographics among participants (e.g., age and educational level). Furthermore, it is conceivable that the psychological orientation underlying the dimensions of social dominance may be more salient or even operate differently among nonstudent adult populations given possible ideological differences across generations. To address this issue, Study 2 replicates Study 1 with a nonstudent, adult sample.

Method: Study 2

Participants and Procedure

A total of 164 participants were recruited using Amazon Mechanical Turk (MTurk) and received monetary compensation for their participation. Two attention check items were included within the questionnaire to ensure that participants were carefully responding to questions. All participants correctly answered the attention check items. The average age of the sample was 34.28 (SD = 10.01). Participants were predominately White (N = 127), followed by self-identifying as Latino (N = 12), Black/African American (N = 11), Asian (N = 9), and multiethnic/multiracial (N = 5). Participants were 64% (N = 105) men, 35% (N = 57) women, and 1% (N = 2) reported as non-gender conforming or “other.” Eighty-six percent (N = 141) reported having attended some college or more, and 50% (N = 82) reported an annual income of US$50,000 or higher. Twenty-five percent (N = 42) identified as Republican, 45% (N = 73) as Democrat, and 30% (N = 49) as independent or “other.” Finally, 48% (N = 79) identified as nonreligious and 52% reported being religious in some capacity (e.g., Christian, Jewish, or Protestant).

Before taking part in the study, participants were informed that their responses were voluntary, anonymous, and that they could quit the study at any time. The entire study took place online. Participants initially answered a series of questions addressing perceptions of authoritarianism, elitism, and social dominance. Following this, participants were presented with the same randomly assigned online news conditions from Study 1. Next, participants answered questions probing emotional responses and social judgments of social groups. Finally, participants answered basic demographic questions (e.g., gender, race). To ensure that participants responded honestly and to reinforce that their answers would not be judged individually, subjects were explicitly given the following instructions multiple times throughout the study: “Please respond to the following statement(s) to the best of your ability. There are no right or wrong answers to any of these statements; we are interested in your honest reactions and opinions.” Finally, participants were presented with a debriefing statement discussing the nature of the study and acknowledging that the online news conditions were created from actual online news article but were fictional.

Moderating Variables

Social dominance

SDO measures (Ho et al., 2015; Pratto et al., 1994) were again used for Study 2, creating two known, psychologically distinct subscales (see Ho et al., 2015) assessing the following: (a) group dominance (α = .91, M = 2.79, SD = 1.52) and (b) anti-egalitarianism (α = .91, M = 5.38, SD = 1.50). The eight-item measure of group dominance included statements such as “Some people are just more worthy than others” and “Some people are just inferior to others.” The six-item measure of anti-egalitarianism included statements such as, “If people were treated more equally, we would have fewer problems in this country” and “We should try to treat one another as equals as much as possible.” Response options ranged from “strongly agree” to “strongly disagree” on a 7-point scale, with higher numbers reflecting greater adherence to SDO.

Dependent Variables

Group emotions

Consistent with Study 1, group-level emotions were assessed with the following single item measures: agitated, offended, ashamed, disgusted, afraid, inspired, proud, hopeful, guilt. Filler items were again included to help conceal the intent of the questions. Participants were asked to rate the extent to which they felt each of the emotions after reading their article, from “not at all” (1) to “very strongly” (7).

Results: Study 2

H1: Social Unrest in the News, Social Dominance, and Negative Affect

H1a–e predicted that news coverage of social unrest, race of the depicted participants, and the social dominance of the reader would interact in predicting negative affective responses to the content. No significant interactions emerged in GLM tests examining H1a–e among this sample of nonstudent, adults. Thus, H1a–e were not supported. Given this, one-way analysis of variance was utilized to examine any potential influence of social unrest news coverage on emotional responses. ANOVAs revealed a main effect of news condition on agitation, F(3, 163) = 3.09; p < .05, $η^{2} = . 06$ , offense, F(3, 163) = 5.90; p < .01, $η^{2} = . 10$ , shame, F(3, 163) = 4.46; p < .01, $η^{2} = . 08$ , disgust, F(3, 163) = 8.44; p < .001, $η^{2} = . 14$ , and fear, F(3, 163) = 4.19; p < .01, $η^{2} = . 07$ .

Specifically, Scheffe post hoc tests revealed that disgust was significantly greater (p < .01) in the White riot condition (M = 3.84, SD = 1.95) than in either the Black (M = 2.13, SD = 1.82) or White (M = 2.12, SD = 1.63) protest conditions. Feelings of taking offense were highest in the White riot condition (M = 3.21, SD = 1.98) and differed significantly from the Black riot (M = 2.02, SD = 1.65, p < .025), Black protest (M = 1.82, SD = 1.34, p < .01), and the White protest conditions (M = 2.05, SD = 1.67, p < .025). Agitation was highest in the White riot condition (M = 3.19, SD = 1.67), with no significant differences across pairs in Scheffe tests. Shame was significantly greater in the White riot condition (M = 3.07, SD = 2.19, p < .01) than in the Black protest condition (M = 1.72, SD = 1.34). Finally, ratings of fear were highest in the White riot condition (M = 3.21, SD = 1.92, p < .025) and differed significantly in Scheffe tests from the Black protest condition (M = 1.95, SD = 1.38).

Research Question: Social Unrest in the News and Positive Affect

The research question explored whether positive affective responses to news were influenced by the type of coverage of social unrest, race of the depicted participants, and the social dominance of the consumer. Univariate GLM revealed a significant three-way interaction between news condition (protest vs. riot), race of the depicted participants (Black vs. White), and the group dominance orientation of SDO in predicting both hope, F(3, 163) = 3.17; p < .05; partial $η^{2} = . 06$ , and guilt, F(3, 163) = 3.00; $η^{2} = . 05$ ; partial $η^{2} = . 06$ . As Figure 5 reveals, as group dominance increased, hope sharply increased for those exposed to the Black and the White riot news story but decreased in the White protest condition, revealing a crossover interaction. With regard to guilt, as group dominance increased, guilt sharply increased in the White riot and White protest conditions, but not the Black riot and protest conditions (see Figure 6). A direct effect of news condition on feelings of pride also emerged, F(3, 163) = 3.85; p < .05, $η^{2} = . 07$ , such that pride was significantly greater (p < .025) in the White protest condition (M = 2.83, SD = 1.79) than in the Black riot condition (M = 1.59, SD = 1.47).

Figure 5.

Three-way interaction between news condition (riot vs. protest), race condition (Black vs. White), and group dominance in predicting hope.

Figure 6.

Three-way interaction between news condition (riot vs. protest), race condition (Black vs. White), and group dominance in predicting guilt.

A significant three-way interaction in predicting hope also emerged between news condition, race of the depicted participants, and the anti-egalitarianism orientation of SDO, F(3, 164) = 3.85; p < .025; partial $η^{2} = . 07$ . As revealed in Figure 7, hope decreased precipitously in the White riot condition and increased in both the Black and White protest conditions as anti-egalitarianism increased.

Figure 7.

Three-way interaction between news condition (riot vs. protest), race condition (Black vs. white) and anti-egalitarianism in predicting hope.

General Discussion

The current set of studies examined the influence of news characterizations consistent with the protest paradigm (e.g., McLeod, 2007) on the intergroup emotions evoked regarding the social actors engaged in civic engagement. Based on theory and empirical evidence in the domain of IET (Mackie & Smith, 2015), it would be expected that news coverage of sensationalized social unrest consistent with the protest paradigm would create a context likely to prompt unfavorable intergroup emotions, that exacerbate unsympathetic intergroup dynamics in society. The results found here indicate that news images of dramatized social unrest can provoke heightened, complex emotional responses beyond mere positive or negative feelings. However, the extent to which these responses vary based on the racialization of the coverage or aspects of psychological group identification appears to be inconsistent across demographic groups. Nonetheless, these data tentatively reveal that the unique psychological dimensions of group dominance orientation can differentially affect how news consumers respond to dramatized and racialized coverage of social unrest.

Study 1: College Students

Among college students, increasing embodiment of the group dominance orientation of SDO, which reflects belief in group-based hierarchy, was predictive of intensified outrage and affront when exposed to Blacks, but not Whites, engaged in aggressive social unrest. This is not a trivial outcome. Research on group-based emotions suggest that distinct behaviors are associated with specific group-level emotions (Maitner et al., 2006) including anger (Yzerbyt et al., 2003), and guilt (Doosje et al., 1998), with subsequent behaviors including attacking and compensatory actions, respectively. Given the efforts of current social movements to curb police violence against Blacks (and non-White racial/ethnic groups in general), this may truly be a grave matter. Of course, neither behaviors nor behavioral intentions were measured in the current study, however, these results suggest that such examinations are warranted in future research.

When it comes to the impact of the anti-egalitarianism dimension of SDO (reflective of opposition to group equality), these data indicate that among college students, increasing embodiment of this psychological orientation is associated with elevated shame and disgust (i.e., avoiding and excluding emotions) when exposed to aggressive displays of social unrest by Whites. Although this result was unexpected, in hindsight, it is not entirely antithetical to existing scholarship. Consistent with research on vicarious shame (e.g., Schmader & Lickel, 2006), it is plausible that exposure to news coverage of this kind (which may be perceived to reinforce an undesirable feature of the group), encouraged distancing motivations in an effort to manage the threat to an identity which subtly supports the exclusion of others from access to resources. In other words, because anti-egalitarianism is fundamentally an endorsement of group exclusivity and social inequality, seeing ingroup members engaged in aggressive efforts to upending economic and structural inequity may produce shame—perhaps in response to the groups own failures or possibly in denial of them. From an IET perspective, then, avoiding or misidentifying would likely emerge from these emotional responses; ultimately delegitimizing the efforts of the social actors but not placing them in physical peril. Certainly, this is an important question to examine in future research.

Next, Study 1 explored if and how exposure to racialized news coverage of social unrest affected college students’ feelings of inspiration, pride, hope, and guilt. Results revealed no effect of either the race of the depicted participants or the social dominance of the audience member on affective responses. However, the level of dramatization of coverage (i.e., riot vs. protest) did influence the emotional response. Specifically, higher levels of inspiration, hope, and pride were reported when exposed to the news coverage of protests (vs. riots). On one hand, this may reflect the social climate of the institution at which these data were collected, which may be predisposed to more favorable views on controlled social protest. On the other hand, it also offers some support for assertions that the dramatization of social unrest in news coverage, as articulated in the protest paradigm research, undermines the ability of social movements or social justice efforts to enact social change.

Study 2: Adult, Nonstudents

For nonstudent adults, the findings were less coherent. The hypothesized interactions between news depictions of social unrest, the race of the depicted participants, and consumers’ SDO did not emerge. It is notable, however, that the White riot condition consistently produced the highest ratings on all negative emotions. Specifically, feelings of agitation, offense, shame, disgust, and fear were significantly greater in the White riot condition compared with other race and protest conditions. This finding is thought-provoking for two primary reasons. First, the fact that exposure to ingroup members (i.e., Whites) involved in antisocial actions (i.e., rioting) prompted heightened feelings of agitation, offense, shame, and disgust may reflect the perception that this behavior places the group in a negative light. Such an interpretation is consistent with empirical evidence in the domain of identity denial (e.g., Branscombe et al., 1999). Identity denial suggests that when ingroup members engage in behaviors that are inconsistent with the perceived values and norms of the ingroup, group members are likely to distance from and harshly evaluate deviant ingroup members, in an effort to protect self and group concept. Unfortunately, without appropriate data to corroborate such an explanation, it remains purely speculative.

Second, the fear response in the White riot condition found in Study 2 differs from Study 1 which revealed fear to be consistently experienced by college student participants when exposed to riot coverage, irrespective of the race of the depicted participants. It is possible that this difference stems, in part, from threats to the environmental realism inherent to the lab-based setting used with the college students (Study 1). At the same time, the heightened fright response to ingroup members engaged in riots (vs. outgroup members), found in Study 2, is difficult to explain in a manner consistent with existing intergroup or media theory. Perhaps the sight of ingroup members rioting provoked concern among these nonstudent White adults that their own communities might be targeted in such acts.

Finally, the extent to which racialized news coverage of social unrest impacted nonstudent adults’ inspiration, pride, hope, and guilt were explored (RQ). Again, findings on the whole were counterintuitive. For example, increasing group dominance was associated with greater feelings of hope when exposed to both Black and White riot coverage and with decreased feelings of hope when exposed to White protest coverage. Alternatively, as anti-egalitarianism increased, feelings of hope decreased in the White riot condition and increased in the Black and White protest conditions. These disparate results suggest that psychological orientation influences emotional responses to news coverage of social unrest, but in largely antithetical ways. To illustrate, given that social dominance is aligned with old-fashioned racism, experiencing hope when exposed to Black riots is perplexing.

Furthermore, the fact that hope increased in both the Black and White protest conditions as anti-egalitarianism increased is also inconsistent with the subtle forms of bias that would be expected for those aligned with this psychological orientation. Results from Study 2 were counter to the stated predictions. However, it may be assumed that affective responses, such as hope, to news images of racialized protest, endorses or reaffirms established stereotypes of Black individuals as criminals (Dixon, 2006) or White liberals as intolerant (Dupree & Fiske, 2019) among audiences demonstrating SDO. Again, without adequate data to substantiate these explanations, this is also speculative; however, continued examination is warranted.

Overarching Implications

Although the anticipated effects of media exposure on audiences are modest, it is certainly the case that problematic media content may exacerbate existing stereotypes and stigmas as well as feelings about one’s ingroup; ultimately influencing interactions between individuals in society. For example, the results that emerged for college participants in terms of expressions of emotions that would be linked with attack (i.e., agitated, offended) versus avoidance (i.e., shame, disgust) behaviors toward the outgroup race (i.e., Black) appeared to exacerbate emotional responses to news coverage of social unrest in a manner that is likely to disproportionately threaten the efforts of race-related collective action movements (e.g., Black Lives Matter).

The difference between Study 1 and Study 2 populations, including the variations in age, geographic location(s), class status, and education levels, may have contributed to the inconsistencies in findings. As stated previously, activating various social categorizations or identities (e.g., age, class) among individuals may result in distinctive emotional experiences (Mackie & Smith, 2015), and this was consistent with these outcomes. Study 2’s sample population (in comparison with Study 1) reflected greater variation in social categories and identities, indicating a less homogeneous group. In other words, similarity across Study 1 participants may have produced more uniform attitudes and behavioral intentions, compared with those found in Study 2’s broader sample. Collectively, the set of findings do not invalidate the overarching conclusions, but rather demonstrate (a) the importance of considering the group-based psychological features of audiences when examining affective responses to dramatized and racialized news coverage of social unrest, and (b) the necessity for continued examination of social unrest, across various audiences and group identities, as each uniquely contributes to human response of media stimuli.

Given that news media’s reach (including print, television, social media, etc.) goes beyond localized settings, audiences’ media exposure, and the influence of this may be of consequence. The transactional relationship between news media content and audiences is increasing and expanding at a rapid pace. Accordingly, we are more connected than ever before and are potentially susceptible to the effects detailed in protest paradigm research. For example, the social unrest among Palestinians and Israelis, protests regarding the treatment of refugees in Indonesia and Greece, and protest marches in the United States and across the globe to raise awareness of cruelty toward and dehumanization of marginalized groups, demonstrate the importance of news media in disseminating critical information about citizens and society. These occurrences underscore the significance of the ways in which media outlets frame protest coverage and the potential influence of this coverage on audiences.

Moreover, when considering that media coverage of radical social protests often treat demonstrators and the issues they represent as trivial or threatening, the result may ultimately delegitimize their efforts to play a role in democratic decision making (McLeod, 2007). The emotions that may result from engagement in media coverage have the potential to spill over into intergroup interactions and affect voting habits, policy implications, and potentially impact the civil liberties of citizens. The research presented here begins the dialogue of news media portrayals of social unrest and potential effects on audiences’ emotional well-being, but the implications of this and its global reach deserve considerable attention.

Considering the wide range of opportunities (e.g., digital news media, social media, and hard copy news) to acquire news addressing collective action, even from a transnational perspective (e.g., social unrest among Palestinians and Israelis, the Dakota Access Pipeline Protest, and protests targeting Brexit), there is a potential likelihood that audiences around the globe may encounter these narratives. The opportunity to thoughtfully engage with news media content featuring collective action may be increasing across society and because of this, neutrality regarding coverage of actors engaged in collective action may benefit readers in their perceptions of those actors. Perhaps when journalists treat a given protest group more positively or aim to objectively represent demonstrators and their circumstance, coverage may provide a cue to audiences that the issues merit attention and consideration. Likewise, news media may also aid in attenuating stereotypes regarding the actors depicted in news coverage by providing more impartial news coverage.

At the same time, this study cannot speak to whether or not consumers would be more likely to select unbiased or issue-based news content addressing collective action or if audiences would be more driven to engage in issue-based news content versus sensationalized news coverage. However, research suggests that salacious news content, although potentially easier to read and understand, may not be more arousing than issue-orientated coverage nor does sensationalized news drive audience consumption (Uribe & Gunter, 2007). Considering this, issue-based news content that depicts narratives of collective action in an unbiased way may potentially benefit the actors involved and support positive intergroup engagement.

This study draws attention to the practice of sensationalizing news coverage of collective action (i.e., the protest paradigm) and the potential impact these practices may have on audience emotions. Thus, drawing attention to the need for careful consideration among decision makers when presenting news stories of collective action and encouraging critical engagement among consumers when thoughtfully reading through material addressing collective action. From the audiences’ perspective, cultural competency and critical engagement may also aid in redirecting the potential impact of the protest paradigm. The evolving, dynamic process of audiences critiquing news media and exercising humility among fellow citizens would be a progressive step forward for individuals and society as a whole.

Conclusion and Limitations

Taken together, the results from Study 1 and Study 2 cautiously suggest that the sensationalized news frames commonly used in coverage of social unrest can provoke distinct group-based affective responses. Additional research will be needed to further flesh out these emotional responses and assess whether the expected behavioral outcomes are, indeed, evoked based on exposure to messages consistent with the protest paradigm. Still, the current data suggests that such outcomes are plausible. If this is borne out by future research, it could mean the difference between harming versus helping different groups in society.

This attempt to investigate the role of intergroup emotions in the context of exposure to news images of racialized social unrest, is not without limitations. Like many studies of intergroup emotions and prejudice, we rely on self-report assessments which may be biased by participants’ inability to accurately report on their own feelings or by their desire to present a more socially desirable response. Furthermore, the use of a generic news story and the experimental setting may have created an excessively artificial environment that, while important for control, limited the responses that might have truly been experienced in society. In addition, research on intergroup emotions reveals that although mere categorization can provoke such group-level emotions, the extent of cognitive and/or emotional identification with the group can moderate these emotional responses and the consequences of categorization. As such, additional research in this context would benefit from incorporating measures of group identification into future tests of such relationships.

In addition, there is precedent to consider the differences between heavy and light media consumers regarding the impact of news media use and emotions evoked from exposure to collective action (McLeod, 1995, 2007). Moving forward, future research should address this issue by measuring media consumption prior to exposure to the experimental manipulation to determine if outcomes differ between heavy versus light media consumers. Despite these limitations, this study offers preliminary insights regarding social unrest and race; a timely and meaningful phenomenon that is interwoven into the fabric of U.S. society.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

David Stamps

Author Biographies

David Stamps (PhD, University of California, Santa Barbara) is currently an assistant professor at the Manship School of Mass Communication at Louisiana State University. His research focuses on representations of marginalized groups in the mass media and the impact of mass media imagery on audience members.

Dana Mastro (PhD, Michigan State University) is professor and vice-chair of the Department of Communication at the University of California, Santa Barbara. Her research documents depictions of Latinos in English and Spanish language U.S. media and empirically examines the range of intergroup and identity-based outcomes associated with exposure to these portrayals.

References

Arpan

L. M.

Baker

Lee

Jung

Lorusso

Smith

(2006). News coverage of social protests and the effects of photographs and prior attitudes. Mass Communication & Society, 9, 1–20.

Atwell Seate

Chien

Mastro

. (2018). Cultivating intergroup emotions: An intergroup threat theory approach. Mass Communication & Society, 21, 178–197. https://doi.org/10.1080/15205436.2017.1381262

Atwell Seate

Mastro

. (2017). Exposure to immigration in the news: The impact of group-level emotions on intergroup behavior. Communication Research, 44, 817–840. https://doi.org/10.1177/0093650215570654

Bassett

J. F.

(2010). The effects of mortality salience and social dominance orientation on attitudes toward illegal immigrants. Social Psychology, 41, 52–55.

Betti

(2016, September 28). The difference between a protest and a riot: When the media misleads people on what a protest is. Odyssey.com [website] Retrieved from: https://www.theodysseyonline.com/the-difference-between-protest-and-riot

Boyle

M. P.

McCluskey

M. R.

McLeod

D. M.

Stein

S. E.

(2005). Newspapers and protest: An examination of protest coverage from 1960 to 1999. Journalism & Mass Communication Quarterly, 82, 638–653.

Boyle

M. P.

McLeod

Armstrong

(2012). Adherence to the protest paradigm: The influence of protest goals and tactics on news coverage in U.S. and international newspapers. International Journal of Press/Politics, 17, 127–144.

Branscombe

N. R.

Ellemers

Spears

Doosje

, (1999). In Ellemers

Spears

Doosje

(Eds.), The context and content of social identity threat. Social identity: Context, commitment, content (pp. 35–58). Oxford: Blackwell.

Brasted

(2005a). Framing protest: The Chicago Tribune and the New York Times during the 1968 democratic convention. Atlantic Journal of Communication, 13, 1–25.

10.

Brasted

(2005b). Protest in the media. Peace Review: A Journal of Social Justice, 17, 383–388.

11.

Caprariello

P. A.

Cuddy

A. J.

Fiske

S. T.

, (2009). Social structure shapes cultural stereotypes and emotions: A causal test of the stereotype content model. Group Processes & Intergroup Relations, 12, 147–155.

12.

Danso

H. A.

Sedlovskaya

Suanda

S. H.

(2007). Perceptions of immigrants: Modifying the attitudes of individuals higher in social dominance orientation. Personality and Social Psychology Bulletin, 33, 1113–1123.

13.

Dixon

T. L.

(2006). Psychological reactions to crime news portrayals of Black criminals: Understanding the moderating roles of prior news viewing and stereotype endorsement. Communication Monographs, 73, 162–187.

14.

Doosje

Branscombe

N. R.

Spears

Manstead

A. S.

(1998). Guilty by association: When one’s group has a negative history. Journal of Personality and Social Psychology, 75, 872–886.

15.

Dupree

C. H.

Fiske

S. T.

(2019). Self-presentation in interracial settings: The competence downshift by White liberals. Journal of Personality and Social Psychology, 117, 579–604.

16.

Heaven

P. C.

(1999). Attitudes toward women’s rights: Relationships with social dominance orientation and political group identities. Sex Roles, 41, 605–614.

17.

Sidanius

Kteily

Sheehy-Skeffington

Pratto

Henkel

Foels

Stewart

(2015). The nature of social dominance orientation: Theorizing and measuring preferences for intergroup inequality using the new SDO Scale. Journal of Personality and Social Psychology, 109, 1003–1028.

18.

Iyer

Schmader

Lickel

(2007). Why individuals protest the perceived transgressions of their country: The role of anger, shame, and guilt. Personality and Social Psychology Bulletin, 33, 572–587. https://doi.org/10.1177/0146167206297402

19.

Leonard

D. J.

Moons

W. G.

Mackie

D. M.

Smith

E. R.

(2011). “We’re mad as hell and we’re not going to take it anymore”: Anger self-stereotyping and collective action. Group Processes & Intergroup Relations, 14, 99–111.

20.

Mackie

D. M.

Devos

Smith

E. R.

(2000). Intergroup emotions: Explaining offensive action tendencies in an intergroup context. Journal of Personality and Social Psychology, 79, 602–616.

21.

Mackie

D. M.

Maitner

Smith

(2009). Intergroup emotion theory. In Nelson

(Ed.), Handbook of prejudice, stereotyping, and discrimination (pp. 285–307). Taylor & Francis.

22.

Mackie

D. M.

Silver

L. A.

Smith

E. R.

(2004). Intergroup emotions: Emotion as an intergroup phenomenon. In Tiedens

L. Z.

Leach

C. W.

(Eds.), The social life of emotions (pp. 227–245). Cambridge University Press.

23.

Mackie

D. M.

Smith

E. R.

(2015). Intergroup emotions. In Mikulincer

Shaver

P. R.

(Eds.), APA handbook of personality and social psychology: Vol. 2. Group processes (pp. 263–293). https://doi.org/10.1037/14342-010

24.

Mackie

D. M.

Smith

E. R.

Ray

D. G.

(2008). Intergroup emotions and intergroup relations. Social and Personality Psychology Compass, 2, 1866–1880.

25.

Maitner

A. T.

Mackie

D. M.

Smith

E. R.

(2006). Evidence for the regulatory function of intergroup emotion: Emotional consequences of implemented or impeded intergroup action tendencies. Journal of Experimental Social Psychology, 42, 720–728.

26.

Maitner

A. T.

Mackie

D. M.

Smith

E. R.

(2007). Antecedents and consequences of satisfaction and guilt following ingroup aggression. Group Processes & Intergroup Relations, 10, 223–237.

27.

McLeod

D. M.

(1995). Communicating deviance: The effects of television news coverage of social protest. Journal of Broadcasting & Electronic Media, 39, 4–19.

28.

McLeod

D. M.

(2007). News coverage and social protest: How the media’s protect paradigm exacerbates social conflict. Journal of Dispute Resolution, 1, 185–194.

29.

Miller

D. A.

Smith

E. R.

Mackie

D. M.

(2004). Effects of intergroup contact and political predispositions on prejudice: Role of intergroup emotions. Group Processes & Intergroup Relations, 7, 221–237.

30.

Moons

W. G.

Leonard

D. J.

Mackie

D. M.

Smith

E. R.

(2009). I feel our pain: Antecedents and consequences of emotional self-stereotyping. Journal of Experimental Psychology, 45, 760–769. https://doi.org/10.1016/j.jesp.3009.04.016

31.

Powell

T. E.

Boomgaarden

H. G.

De Swert

de Vreese

C. H.

(2015). A clearer picture: The contribution of visuals and text to framing effects. Journal of Communication, 65, 997–1017.

32.

Pratto

Sidanius

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, 741–763.

33.

Ramasubramanian

(2010). Television viewing, racial attitudes, and policy preferences: Exploring the role of social identity and intergroup emotions in influencing support for affirmative action. Communication Monographs, 77, 102–120.

34.

Ray

D. G.

Mackie

D. M.

Rydell

R. J.

Smith

E. R.

(2008). Changing categorization of self can change emotions about outgroups. Journal of Experimental Social Psychology, 44, 1210–1213.

35.

Schmader

Lickel

(2006). Stigma and shame: Emotional responses to the stereotypic actions of one’s ethnic ingroup. In Levin

Van Laar

(Eds.), Stigma and group inequality: Social psychological perspectives (pp. 262–280). Lawrence Erlbaum.

36.

Skitka

L. J.

Bauman

C. W.

Mullen

(2004). Political tolerance and coming to psychological closure following the September 11, 2001, terrorist attacks: An integrative approach. Personality and Social Psychology Bulletin, 30, 743–756.

37.

Smith

E. R.

Seger

C. R.

Mackie

D. M.

(2007). Can emotions be truly group level? Evidence for four conceptual criteria. Journal of Personality and Social Psychology, 93, 431–446.

38.

Uribe

Gunter

(2007). Are sensational news stories more likely to trigger viewers’ emotions than non-sensational news stories? A content analysis of British TV news. European Journal of Communication, 22, 207–228.

39.

Weaver

D. A.

Scacco

J. M.

, (2013). Revisiting the protest paradigm: The Tea Party as filtered through prime-time cable news. The International Journal of Press/Politics, 18, 61-84.

40.

Whitley

B. E.

Ægisdόttir

, (2000). The gender belief system, authoritarianism, social dominance orientation, and heterosexuals’ attitudes toward lesbian and gay men. Sex Roles, 42, 947–967.

41.

Yzerbyt

Dumont

Wigboldus

Gordijn

(2003). I feel for us: The impact of categorization and identification on emotions and action tendencies. British Journal of Social Psychology, 42(4), 533–549.