The Emotional Power of Words and Visuals
in Terrorism News Coverage
Master’s Thesis
Graduate School of Communication
Master’s programme Communication Science
University of Amsterdam, Netherlands
Linda Henke
Student number: 12033790 | E-Mail: linda.henke@berlin.de
Supervisor: Laura Jacobs | Date: 25.06.2019
The Emotional Power of Words and Visuals in Terrorism News Coverage
Abstract
This study focuses on the effects of linguistic and visual decisions in terrorism news on perceived terrorist threat and negative emotions with special attention on fear and anger. The mere terrorist labelling, as well as visuals differing in their threat level, are subjects of this study. The current study hypothesizes that terrorist labelling will positively affect
negative emotions in general, fear more specifically and perceived terrorist threat. Moreover, it is expected that terrorist labelling a) combined with a neutral visual and b) combined with a threatening visual will increase this effect. Within an experimental study among a German sample (N = 271), the labelling of an attack and the type of visual imagery has been manipulated. Results refute the hypotheses at several points. While the news story without labelling significantly increases negative emotions, non-significant results are found for perceived terrorist threat. Multi-modal presentations of news increase negative emotions but for negative emotions in general only the threatening visual matters. Surprisingly, in contrast to fear, anger significantly increases as a response to news exposure. For anger, the neutral and the threatening visual enhances effects compared to no visual at all. Results of this study broaden the scope of research on news production and emotional responses and clearly underline that journalistic decisions with regard to terrorist attacks have meaningful consequences.
Introduction
Terrorism is one of the most central global risks and does not stop at German borders. While Germany’s experience with terrorism goes back far, the recent wave of ideologically motivated terrorism occurring in Western countries found entrance into every area of German society. Latest with the assassination on a Berlin Christmas market in December 2016,
terrorism is amongst the most frequent perceived threats for German citizens according to recent public opinion data (e.g., Institut für Politische Wissenschaft, 2018). The news media plays a significant role with regard to terrorism and deserves special attention. As described by Morin (2016), news and terrorism are interrelated through a “symbiotic relationship” (p. 988). On the one hand, terrorist attacks are by definition exceptional and shocking events which lead them to be widely covered in the news media. From a journalistic perspective, terrorist attacks are highly newsworthy events due to the following internal news values (Lerner, Gonzalez, Small & Fischhoff, 2003): their inherent negativity, their relevance, the conflict value as well as the element of surprise and the sudden shock of a terrorist attack (Cho et al., 2003). Terrorism, on the other hand, depends on the orchestration of terrorist attacks by means of the media to spread its intimidating message (Shoshani & Slone, 2008). Consequently, the considerable challenge that news media face is twofold: it has a democratic duty to cover terrorism as well as reporting in a responsible way that does not benefit
terrorists’ strategy.
Since terrorist attacks physically impact a comparably small group of citizens directly, most citizens rely on the news to make sense of the event. As extensively scrutinized by earlier research, the news media shapes citizen’s perception of reality (e.g., Morin, 2016; Vasterman, 2005; Yang & Chen, 2018). In line with this, the level of emotions has been found to partly depend on the information conveyed and the way terrorism is reported on in the news (Von Sikorski, Schmuck, Matthes & Binder, 2017). More specifically, the impact of
journalists’ linguistic and visual decisions on negative emotions and perceived terrorist threat has been confirmed in prior research. Thus, news frames and emotionally laden information boost negative emotions (Lecheler, Schuck & de Vreese, 2013; Nabi, 2002). Moreover, terrorist labelling of an attack and negatively valenced terrorism news strengthen perceived terrorist threat (Atwell Seate & Mastro, 2017). Yet, news frames and emotion-induced news have been the main subject of investigations (Cho et al., 2003), while the impact of the sober terrorism label has not received much attention. In reference to emotions, it is well-known that they strongly impact political behaviour (Kühne & Schemer, 2015; Lecheler, Schuck & de Vreese, 2013). While prior scholarly attention shows a tendency to focus on emotions in their mediating role, the direct relationship between terrorism news and emotions lament a lack of empirical attention (Atwell Seate & Mastro, 2017; Lecheler, Schuck & de Vreese, 2013; Uenal, 2016). The study at hand sets out to fill this gap and aims to investigate the effects of terrorist labelling of an attack in a news story on negative emotions and perceived terrorist threat. Hence, it integrates into a unique approach and tests whether emotions are elicited by the mere labelling without consciously induced emotional appeals in a news article.
The following study also addresses the role of visual imagery for inducing negative emotions and perceived terrorist threat based on the theoretical foundation of the dual-code-theory (Abraham & Appiah, 2006). The relevance of visuals becomes clear when considering the fact that news increasingly applies multi-modal ways of information presentation (Geise & Baden, 2014). Existing research offers evidence for the relevance of visuals for the
emergence of emotions (Iyer, Webster, Hornsey & Vanman, 2014; Morin, 2016; Shoshani & Slone, 2008). Particularly when combined, terminology and visual imagery evoke strong emotional reactions (Powell et al., 2015) as together they allow for better comprehension and information recalling (Abraham & Appiah, 2006). Accordingly, it is hypothesized that
negative emotions and perceived terrorist threat increase when terrorist labelling is
supplemented with a neutral visual and even more so when supplemented with a threatening visual in a news article.
This study adds to a current trend in research on emotions and news coverage (Lecheler, Schuck & de Vreese, 2013) and anticipates that the distinct emotions fear and anger yield diverging effects. Within the framework of the Appraisal Theory of emotions (Roseman, 1991), the examination of the so-called determinants human agency and uncertainty lead to the expectation that fear will be induced stronger than anger. By integrating the perception of risk as an additional foundation, the theoretical framework provides a comprehensive contextualization of some of the most relevant aspects that support this expectation. In applying the Appraisal Theory to the concept of terrorism, this article expands its field of application. To put the hypotheses to an empirical test, an
online-embedded experiment was conducted with a total sample of 271 German participants. Due to its experimental design, the study contributes to the external validity of existing work in this field.
To start, the first theory section provides a short excursus about terrorism which is crucial to understand its semantical power as a base for this study. In the subsequent sections of the theoretical framework, it will be elaborated on the theoretical definitions underlying the concepts discrete emotions and perceived terrorist threat. In an attempt to integrate the role of the news media at all points, the concepts will be contextualized with findings from prior research that suggest relevant aspects resulting in the formulation of hypotheses.
Terrorism
The American-based Global Terrorism Database (GTB) inclusively defines terrorism as “actual use of illegal force and violence by a non-state actor to attain a political, economic,
religious, or social goal through fear, coercion, or intimidation.” (GTB, 2018).Terrorism is mostly grounded in political or ideological beliefs1. However, its unique mode of expression, its method and “modus operandi” (Armborst, 2010, p. 422) constitute its powerful
component. Hence, terrorism is a form of violence which aims at attacking a whole society (mostly within country borders) and its core values (safety, freedom, peace) on the long-term by physically attacking a group of people (Morin, 2016). The core strategy of terrorism is to imprint terror into the collective consciousness. Consequently, led by a scheme of uncertainty, terrorist attacks often occur at wilfully chosen public places with large crowds, such as train stations (e.g., Madrid, 2004, London, 2005), religious places of worship (e.g., Sri Lanka, 2019) or other public places (e.g., Paris, 2015, Nizza, 2016, Berlin, 2016). This study deals with the form of terrorist attacks as they occur in Western societies.
As preliminary mentioned, the media plays a significant role in terrorism’s echo. With this in mind, the significance of examining journalistic decisions more closely is sufficiently emphasized. How these decisions might impact negative emotions and the perception of terrorist threat, will be elaborated in the following sections.
Discrete Emotions
Emotions are defined as an immediate and sudden response to external stimuli
(Marcus, 2000). They are “generally short-lived [and] intense” (Nabi, 2002, pp. 289-290) and message-related reactions. As such, emotions are opposed to moods and traits. While longer-lasting moods extend over context and time (Atwell Seate & Mastro, 2017; Lecheler, Schuck & de Vreese, 2013), traits are defined as personal characteristics (Nabi, 2002). Among the psychological field, a broad variety of understandings of emotions exists (Kühne & Schemer, 2015; Lecheler, Schuck & de Vreese, 2013; Marcus, 2000). The concept of discrete emotions which is organized along the overarching line of valence (i.e., whether emotions are generally
negative or positive) applies best to this study. Widely used in prior research, it stands out due to its clear distinction between positive and negative emotions (Ottati, Steenbergen & Riggle, 1992)2. This study focuses on negative emotions as there is no doubt that terrorism is
consensually perceived as a negative event that by nature primarily trigger negative emotions. When shifting the focus on which aspects of news evoke emotional responses, the prominence of emotions’ mediating role between news and several political cognitive and affective reactions becomes apparent (Brouard, Vasilopoulos & Foucault, 2018; Kühne & Schemer, 2015). In this sense, literature referring to mediating emotional effects highlights that the pure content of news should not be underestimated (Fahmy, Cho, Wanta & Song, 2006). Lecheler, Schuck and de Vreese (2013) indicate that exposure to news frames, varying in valence, evokes negative and positive emotions respectively, which mediate subsequent changes in political opinion. In a similar vein, the use of anger- or sadness-induced news frames elicit corresponding emotions, which in turn influence opinion formation and information processing (Kühne & Schemer, 2015).
All in all, the current state of the literature regarding news and emotions can be summarized as follows: first, it is commonly accepted that the content of news impacts negative emotions, but previous research has drawn on frames and emotion-induced aspects in news and mainly disregarded sober terminologies. Second, the considerable role of emotions is emphasized since they influence meaningful political processes, but the direct effect of news on emotions remains to be further investigated.
Perceived Terrorist Threat
Following Doosje and colleagues (2009), the definition of terrorist threat as “perceived danger of becoming a victim of a terrorist attack” (p. 209) underlies this study. The present article follows the current trend of expanding types of threats beyond those
conceptualized in the Integrated Threat Theory (ITT) (Stephan & Stephan, 1996; 2013). The ITT is a prominent threat theory that categorizes intergroup threats into realistic and symbolic threats (Stephan & Stephan, 1996; 2013). Potentially, perceived terrorist threat could be considered a subcategory of realistic threats since those indeed include concerns about physical safety (Obaidi, Kunst, Kteily, Sidanius & Thomsen, 2018).
Nonetheless, this study chose to treat terrorist threat (also categorized as safety threat by Uenal, 2016) as a distinct concept from intergroup threats. Ascribing terrorist threat to the classical categorization (realistic/symbolic) is not fully self-evident when recalling terrorism’s strategy (cf. supra). Obviously, physical violence is a primary characteristic of terrorist
attacks, but the psychological long-term aim to threaten the whole society cannot be denied. This notion is supported by a study conducted by Uenal (2016) who has found that terrorist threat is empirically distinct from realistic threat. Therefore, an accurate investigation of terrorist threat requires strict isolation.
By accepting that terrorist attacks go beyond primarily physical repercussions, it seems reasonable to distinguish between terrorist threat on the societal and personal level as argued by Onraet and Van Hiel (2013). According to them, terrorist threat on the personal level occurs when individuals are afraid they could personally become a victim of an attack, whereas on the societal level people fear an attack on the country and its symbolic values (Onraet & Van Hiel, 2013).
Regardless of the conceptional distinction, terrorism news increases the perception of terrorist threat (Atwell Seate & Mastro, 2017; Obaidi, Kunst, Kteily, Sidanius & Thomsen, 2018; Slone, 2000). More specifically, Gadarian (2010) has shown that a threatening manner of reporting on terrorism news as well as scary visuals lead to increased levels of perceived terrorist threat.
Terminology and Visual Imagery in News
To investigate further which specific aspects of news influence negative emotions and the perception of terrorist threat, this study specifically turns on terminology and visual imagery.
From an empirical point of view, the high significance of words used in news
coverage is not a new insight, as it, for instance, constitutes the base for framing theories. In news coverage of violent events, terrorism is a very prominent label (Morin, 2016) and used almost inflationary. It has been found that readers made unconscious connections between shootings and Islamic terrorism, although no explicit reference to Muslims or Islam was present (Morin, 2016). This finding exemplifies the power of association provoked by
terminology and indicates that individuals highly rely on pre-existing frames and knowledge. Disentangling concrete connotations with the terrorism-label is not an easy undertaking since it is continuously re-defined (Cooper, 2001). Still, this study assumes that merely the use of the terrorism-label triggers a semantical net of meanings which already prevails in people’s minds in times of continuous terroristic attacks in Western societies (Shoshani & Slone, 2008; Morin, 2016). More specifically, news coverage applying terrorist labelling may plant the fact that terrorists attack the whole nation into people’s consciousness by building on existing meta-narratives and connotations (Morin, 2016). This process is called “terrorism’s
mechanism of self-empowerment via the media” (Shoshani & Slone, 2008, p. 627). Given the recent uplift of terroristic attacks and the high significance of language, it can be assumed that the terrorist label evokes strong associations. This leads to the first hypothesis:
H1: Exposure to terrorist labelling of an attack in a news story will induce a) stronger
negative emotions and b) a stronger perception of terrorist threat than exposure to a news story without labelling.
One would expect that audio-visual news is more inclined to evoke emotional responses than print news due to the fact that characteristics inherent in television news are already more emotional in nature (for instance, live broadcasting, real-time reactions by individuals/politicians on location). Previous research shows a preference of assessing effects of television news (Nellis & Savage, 2012; Newhagen, 1998; Shoshani & Slone, 2008), which results in a lack of research on terrorism news in print and online news.
This void is addressed in this study for several reasons. The use of visuals is not exclusively reserved for television news; in contrast, also written news show an extended use of visual content (Fahmy, Cho, Wanta & Song, 2006). As aforementioned, the significant role of visual imagery rapidly gains importance since news is edited in increasingly multimodal ways (Geise & Baden, 2014; Powell, Boomgaarden, De Swert & de Vreese, 2015; Powell, Boomgaarden, De Swert & de Vreese, 2018). Moreover, written news coverage catches up with television news due to digital means (Powell, Boomgaarden, De Swert & de Vreese, 2018).
A noticeable amount of research has confirmed visual’s power to engender emotional responses (Iyer, Webster, Hornsey & Vanman, 2014; Morin, 2016; Shoshani & Slone, 2008). A study conducted by Powell and colleagues (2015) has revealed that exposure to images induces stronger framing effects compared to exposure to text solely. Furthermore, the positive effects of images on the comprehension of information and the ability to recall information in comparison to verbal information are commonly accepted (Abraham & Appiah, 2006).
To the best of my knowledge, however, most research has explored the influence of visual imagery and words separately (Geise & Baden, 2014; Powell, Boomgaarden, De Swert & de Vreese, 2018). When exposing participants to text and image in combination, mediating effects of emotions have been pointed out: anger and fear mediate effects of news exposure
on political behavioural intentions (Powell, et al., 2015). Combined in multi-modal presentations, linguistic and visual content merges. The dual-code-theory (Abraham & Appiah, 2006) predicts an enforced semantical interaction of words and images which is expected to result in strengthened effects on cognitive and affective processes. Based on this notion, this study argues that terrorist labelling combined with visual imagery enforces the perception of terrorist threat and elicits even stronger negative emotions than the terrorist labelling solely.
Apart from the expectation that the mere presence of visual imagery influences discrete negative emotions and perceived terrorist threat, it is plausible that threatening visual content reinforces this effect. In line with the positive effect on memory, visuals’ ability to communicate “details of an event in succinct and captivating ways” (Iyer, Webster, Hornsey & Vanman, 2014, p. 512) is grounded in the interplay with observer’s associations.
Associations with personal vulnerability (Huddy, Feldman, Capelos & Provost, 2002) and mortality salience (Landau et al., 2004) result in increased evaluations of threat presence in news content. The depiction of scenes that remind individuals to direct and indirect
experiences with crime and victimization reinforce negative emotions and threat perception (Gadarian, 2010). Ergo, the following expectations are tested:
H2: The effect of exposure to terrorist labelling of an attack in a news story on
negative emotions increases when a) combined with a neutral visual and b) is strongest when combined with a threatening visual.
H3: The effect of exposure to terrorist labelling of an attack in a news story on
perceived terrorist threat increases when a) combined with a neutral visual and b) is strongest when combined with a threatening visual.
The Distinction between Fear and Anger
Next to the expected effect on negative emotions in general, this study hypothesizes a stronger effect on fear than on anger. At least three reasons justify the decision for the choice of these emotions. First, and most noteworthy, terrorism’s main intention is to generate fear (Von Sikorski, Schmuck, Matthes & Binder, 2017). Second, both anger and fear are two of the main emotions people experience following terrorist attacks (Vasilopoulos, Marcus & Foucault, 2018). Hence, a study conducted after the 2003 Madrid attacks has found that anger and fear were the prime emotions that citizens experienced (Conejero & Etxebarria, 2007). Third, although both are negatively valenced emotions, anger and fear seem to differ in their core, which has granted them high theoretical and empirical attention in previous research (e.g., Lecheler, Schuck & de Vreese, 2013; Lerner & Keltner, 2001; Newhagen, 1998).
In order to examine anger and fear caused by terrorism news, the Appraisal theory is taken into consideration (Roseman, 1991; Smith & Ellsworth, 1985; Watson & Spence, 2007). The theory bridges cognitive and affective processes (Nabi, 1999; 2002) and is founded on the core that evaluations and interpretations of events are more decisive than the events themselves (Roseman, 1991)3. Based on this understanding, individuals evaluate external stimuli unconsciously along different cognitive dimensions or so-called determinants (Smith & Ellsworth, 1985; Watson & Spence, 2007). The scientific aim of considering these determinants is to predict which emotions develop in certain situations (Iyer, Webster, Hornsey & Vanman, 2014)4. This said, the determinants human agency and uncertainty will be elaborated and contextualized with terrorism in the following.
Human agency refers to assigning responsibility to whom or what has control over an event, which may be the circumstances (such as ecological disasters), oneself or someone else (Roseman, 1991; Smith & Ellsworth, 1985). Generally, both anger and fear are elicited when control over a negative event is assigned to someone else (Nerb & Spada, 2001), which
underlines H1. However, fear arises stronger than anger when control is attributed to concrete other persons (Roseman, 1991). Additionally, the aspect of controllability is inherent in assigning human agency. In absolute contrast to anger, fear arises when individuals feel a lack of power about an event (Nabi, 1999; Turner, 2007) and perceive it as personally
uncontrollable (Lerner & Keltner, 2001; Smith & Ellsworth, 1985). In the context of terrorist attacks, it is self-evident that the responsibility for a terrorist attack is attributed to others. Terrorist organizations do not leave much room for doubt since it is their aim to be held responsible for their violent acts. Plus, terrorist attacks do not lie in ordinary people’s control (as perpetrators decide to commit a crime). Besides, the perception of an event as a threat to the own feeling of safety contributes to the emergence of fear, which is unquestionably the case with terrorist attacks as they pose a threat on the physical and psychological level (Nabi, 2002). Following these insights, terrorism news is expected to impact fear more strongly than anger.
Turning to the determinant uncertainty conceptualized in the Appraisal Theory, it is described as the perceived probability of an event (Roseman, 1991; Watson & Spence, 2007). As empirically verified (Roseman, 1991), uncertainty and, therefore, negative assessments toward the likelihood of a negative event boost fear, whereas certainty boosts anger5. With terrorist attacks, it is a unique phenomenon that people perceive the probability as higher than it really is6.
In this sense, the concept of risk perception seems highly relevant in this context. Exposure to terrorism news is positively connected with the level of the perceived risk of terrorism (Nellis & Savage, 2012). A considerable corpus of literature emphasizes the role of risk perception for the evaluation of events as it meaningfully interacts with emotions
Pessimistic risk assessments go hand in hand with fear, meaning that a high level of perceived risk positively impacts fear (Lerner & Keltner, 2001).
In addition, it is worth paying attention to differences between anger and fear
identified beyond the Appraisal Theory. In that sense, it is relevant to note that arguments and heuristic schemata are of strong importance for anger, while factual arguments have less relevance for inducing fear. This suggests that fear relies more on heuristic associations, which relates to the intuitive power of visuals and terminologies7. These theoretical perspectives are formulated into the final hypothesis:
H4: The effect of exposure to terrorist labelling of an attack in a news story induces
fear stronger than anger a) compared to exposure to a news story without labelling, b) when combined with a neutral visual and c) the most when combined with a
threatening visual.
Method Design and Participants
A survey-embedded online experiment was conducted. The experiment was between-subjects and was arranged in a 2 (terrorist labelling vs. no labelling) x 3 (no visual vs. neutral visual vs. threatening visual) factorial design. The experimental design is illustrated in Table 1. Since the study aimed to investigate whether the effect of terrorist labelling on negative emotions increases when adding visuals, only four conditions were used. Ideally, the main effect of visuals without manipulation of terminology would have been tested, but a limited number of participants restricted the feasibility. Participants were randomly assigned to one of the conditions.
Table 1. Design Experiment Terminology Visual Imagery
No visual Neutral visual Threatening visual
Terroristic labelling
Condition A (n=64) Condition C (n=62) Condition D (n=63)
No labelling Condition B (n=62)
Snowball-sampling techniques8 and the online research platform SurveyCircle were used to recruit participants. SurveyCircle enables researchers to recruit external and unknown participants. The study was conducted in German among a German-speaking sample to
exclude a lack of effects due to linguistic barriers. The final sample contained 271 participants aged 18 – 84 and consisted of 60.9% females (n=165) and one participant defining its gender as “non-binary”, Mage = 35.4, SDage = 15.81). The sample was skewed to a political left-wing orientation on a scale from 0 (left) to 10 (right) (M = 3.65, SD = 1.74). In the last
parliamentary election in 2017, the majority voted for the German green party Bündnis90/DIE GRÜNEN (n=80) while the social labour party Sozialdemokratische Partei Deutschlands holds second position (n=55). The vast majority was born in Germany (93%), as well as their parents (85%).
Randomization checks, conducted with chi-square tests and analysis of variance (ANOVA) revealed successful randomization across the four experimental conditions with regard to key socio-demographic characteristics such as gender, χ2 (8, N = 271) = 5.72, p = .679, age, F(3, 268) = 0.17, p = .914 and education (clustered into 2 groups), χ2 (3, N = 269) = 5.04, p = .169.
Procedure
Ethical permission was obtained to conduct the experiment. The questionnaire was distributed via mail, social media platforms and SurveyCircle between May 17 until May 30, 2019. Participants were told that the study investigated news coverage about sensitive topics. Information about the purpose of the study was consciously kept to a minimum to reduce the risk of sensitization and to maximize the success of the manipulation. Moreover, they were informed that they would read a newspaper article which might influence their mood in a way. After explicitly consenting to participate, participants completed a set of questions about their demographics. Afterwards, they were presented with one of the newspaper articles. A timer forced participants to read the article for at least 45 seconds. Manipulation checks assessed whether participants perceived each manipulation as intended. Generally, there may be ethical objections to evoking negative emotions and the perception of threat (Roseman, 1991). Accordingly, a detailed debriefing on the last page of the questionnaire informed participants in detail about the real goal of the study and they were told the news story was fictional. The debriefings differed concurrent with each condition. By following a link, participants could read the original article.
Stimuli
News Stories and Labelling
Print news was tested in the form of online news coverage in this study9. An online news story of approximately 350 words published by Spiegel Online in 2015 was downloaded from the official Spiegel Online website and modified according to the studies’ empirical needs. This approach increased authenticity and, hence, the experiment’s external validity. The original article deals with a terrorist attack the German police was able to prevent in a small German town. Since participants would be distracted from the treatment material when
exposing them to an article about an attack that actually happened, this version provided a balance between internal and external validity. When prevented, attacks receive considerably less media attention, which reduces the likelihood of prior knowledge. Still, referring to a prevented attack may be sufficiently worrying for participants and may still induce emotional responses and perceived terrorist threat. Minor changes beyond the experimental
manipulations in both versions of the article were necessary. Taking geographical proximity as news value into consideration (Boukes & Vliegenthart, 2017; O’neill & Harcup, 2016), the external validity and the relevance of the article were improved by changing the publication date (to 2019) and the location (to the closest big city Frankfurt am Main). To minimize any other influences, references to the foreign background of the suspects and religious-motivated terrorism were deleted. The experimental conditions only differed with regard to the stimulus to avoid confounding influences. Terminology (‘terroristic attack’ / ‘attack’) was adopted throughout the whole article. The description of the culprits was modified from ‘attackers’ to ‘terrorists’ and vice versa.
Visuals
Both visuals were previously published by Spiegel Online. The neutral visual
condition shows a photo of Frankfurt am Main’s skyline, which does not evoke any positive or negative associations and can be considered neutral. According to theoretical rationale, the visual in the threatening visual condition showed several armed police officers in action, which are assumed to evoke associations with personal vulnerability and relate to experiences with crime and victimization.
Pilot Study
A pilot study tested the stimulus materials among a sample of N = 39. Considering that respondents participating in the pilot study were sensitized, the pilot study was conducted
in English among English-speakers. The point of reaching the possible maximum of a German sample size for the main study was preferred over language unity in the pilot and main study. It was found that the relationship between the experimental groups (terrorist labelling / without terrorist labelling) was significant, X2 (1, N = 39) = 13.75, p < .001, φ = 0.6. Results are presented in Table 2. To eliminate any doubt, the newspaper articles were supplemented with two more sentences referring to a terrorist attack/attack. Respondents perceived the threatening visual (M = 81.1, SD = 19.66) as significantly more threatening than the neutral visual (M = 7.85, SD = 13.72), t(38) = 19.95, p < .001, d = 3.2. Thus, both
manipulations were perceived as intended. Refer to Appendices A and B for all stimulus materials and operationalizations of measurements.
Table 2. Results of Terminology Manipulation in the Pilot Study
Response Terrorist Labelling No Labelling Total
n % n % n %
Yes 16 76.2 3 23.8 21 100
No 5 16.7 15 83.3 18 100
Total 21 18 39 100
Note. Question: Did the newspaper article explicitly refer to a terrorist attack (e.g. through
the use of the label ‘terrorist’)?
Measurements
Negative Emotions. Discrete negative emotions were measured by asking participants
to indicate the extent to which they experience various emotional states after reading the newspaper article. Items were measured on an 8-point scale (0 = Not at all, 7 = Extremely
strong). A middle point was avoided to force participants to indicate a direction. The items
.94) were based on distinct scales validated by Iyer, Webster, Hornsey, Vanman (2014). For testing negative emotions in general, anger and fear were summarized (α = .91).
Principal components factor analysis revealed that only one component had an Eigenvalue above 1.00. According to the theoretical rationale, a clear distinction between anger and fear was crucial. A two-factor solution within a principal components factor analysis revealed that the three items measuring anger loaded strongly on the first factor (> .82) and the three items measuring fear loaded on the second factor (> .74). Since some extent of correlation between both negatively valenced emotions can be assumed, oblimin rotation strategy was applied to distinguish between both emotions. Thus, two scales were created measuring anger (α = .89) and fear (α = .86) separately in order to test H4.
Perceived Terrorist Threat. Following the distinction between threat on the societal
and personal level, three items assessed each (based on Cohrs, Kielmann, Maes & Moschner, 2005). Participants had to indicate the extent to which they agree with these statements on a 7-point scale (ranging from Strongly Disagree (1) to Strongly Agree (7)) after reading the newspaper article. The following items were included to measure terrorist threat on the societal level: (1) “Terrorism threatens our society.”; (2) “I feel that the daily life in our country is affected by possible terrorist actions.”; (3) “There is a good chance that a terrorist attack will happen in our country in the near future.” (α = .86). These items measured terrorist threat on the personal level: (4) “I feel that my everyday life is affected by possible terrorist actions”; (5) “I feel personally threatened by possible terrorist actions”; (6) “There is a
realistic chance that I myself or one of my relatives will become a victim of a terrorist attack.” (α = .85) (Onraet & Van Hiel, 2013). Even though validated as two distinct scales, principal components analysis showed that all items loaded on one component with an Eigenvalue greater than 1.00. For the sake of this study, analyses were conducted with two distinct scales as well as with one scale summarizing all items. Results did not considerably differ.
Moreover, the theoretical framework does not require the strict distinction between terrorist threat on the societal and personal level. Hence, one scale measuring terrorist threat was created (α = .84).
Results
The analyses were conducted with and without variables that could possibly influence the outcomes, resulting in identical results10. Table 3 presents descriptive statistics of the dependent variables.
Table 3. Descriptive Statistics of Conditions on Dependent Variables.
A B C D
M (SD) M (SD) M (SD) M (SD)
Negative Emotions 1.85 (1.46) 2.47 (1.94) 2.40 (1.43) 2.64 (1.74) Perceived Terrorist Threat 4.55 (1.21) 4.41 (1.21) 4.24 (0.97) 4.10 (1.14)
Anger 1.76 (1.52) 2.58 (2.12) 2.49 (1.85) 2.65 (2.12)
Fear 1.95 (1.62) 2.37 (1.85) 2.30 (1.44) 2.63 (1.72)
N 64 62 62 63
Note. A = Terrorist labelling; B = No labelling; C = Terrorist labelling + Neutral Visual; D =
Terrorist Labelling + Threatening Visual
Manipulation Checks
Two manipulation checks tested whether participants picked up on the manipulation as intended. Chi-square tests revealed that 81% of the participants perceived the manipulation correctly, X2 (1, N = 251) = 62.43, p < .001, φ = 0.5. Nevertheless, it is crucial to notice that the majority of those who did not answer correctly was exposed to the article without
main effects were conducted including and excluding those who answered the manipulation check incorrectly. As the results did not indicate considerable differences, these participants were not excluded in further analyses. However, all findings related to the labelling have to be taken with caution. A second manipulation check assessed whether participants perceived the visuals as threatening/non-threatening. A t-test showed a significant difference between the neutral visual (M = 19.61, SD = 22.56) and the threatening visual (M = 46.7, SD = 26.55) on the level of threat perception, t(123) = -6.14, p < .001, d = 1.1. Hence, manipulation was successful.
Figure 1. The Effects of Terrorist Labelling on Negative Emotions and Perceived Terrorist
Threat.
Note. Significance: * p < .05.
Terminology
To test whether terrorist labelling of an attack in a news story induces a) stronger negative emotions and b) a stronger perception of terrorist threat than when there is no labelling mentioned (H1), independent samples t-tests were conducted. For H1a, it should be noted that the assumption of equal variances in the population has been violated, Levene’s F
1,85 4,55 2,47 4,41 0 1 2 3 4 5 6 7
a) Negative Emotions b) Perceived Terrorist Threat
A = Terrorist Labelling B = No Labelling
(124, 113) = 10.15, p = .002. The results indicate the opposite direction than expected with significantly stronger negative emotions in the condition without labelling (M = 2.47, SD = 1.94) than in the terrorist labelling condition (M = 1.85, SD = 1.46) of small to medium effect size, t(113.55) = -2.03, p = .045, d = 0.38.
Non-significant results were found for the difference between the labelling conditions on the perception of terrorist threat, t(124) = 0.65, p = .516. For H1b, the assumption of equal variances holds, Levene’s F(124, 123,87) = 0.03, p = .874. Still, the direction of differences revealed that participants exposed to the terrorist labelling (M = 4.55, SD = 1.21) perceived slightly stronger terrorist threat than participants exposed to the article without labelling (M = 4.41, SD = 1.21).
Overall, findings do not support H1a and H1b. It is worth mentioning that even though there was no significant effect on perceived terrorist threat, in general, the level of threat was higher than the level of negative emotions. Figure 1 shows that also in the condition without
labelling, negative emotions were experienced to a relatively small extent compared to perceived terrorist threat.
Visuals
To test the hypotheses whether the effect of the terrorist labelling of an attack in a news story on negative emotions increases when H2a combined with a neutral visual and H2b is strongest when combined with a threatening visual, analyses of variances (ANOVA) were conducted. The condition without terrorist labelling was disregarded according to the
hypotheses. A significant main effect of rather small effect size for the conditions (terrorist labelling and combined with neutral visual and threatening visual) on negative emotions was found, F(2, 186) = 4.28, p = .015, ηp2 = .04.
Figure 2. The Effects of Newspaper Articles Varying in Visuals on Negative Emotions and
Perceived Terrorist Threat.
Note. Significance: * p < .05.
Simple contrasts revealed that the neutral visual (M = 2.4, SD = 1.43) did not significantly increase negative emotions compared to the terrorist labelling without visual (M = 1.85, SD = 1.46) (H2a). Also, simple contrasts showed that there is no significant increase of negative emotions in the threatening visual condition (M = 2.64, SD = 1.74) compared to the neutral visual condition (M = 2.4, SD = 1.43), F = 0.72 (H2b). To further investigate the cause of the significant main effect, simple contrasts revealed that there was a significant increase of negative emotions in the threatening visual condition (M = 2.64, SD = 1.74) compared to the terrorist labelling condition without visual (M = 1.85, SD = 1.46), p = .005, 95 % CI [0.24, 1.33]. Consequently, H2a is rejected, while H2b is confirmed.
Referring to H3a and H3b, which hypothesize the same effect on perceived terrorist threat as on negative emotions in combination with visuals, no significant effects for visual were found on perceived terrorist threat, F(2, 186) = 2.65, p = .073. Simple contrasts showed that the neutral visual (M = 4.24, SD = 0.97) did not significantly increase the perception of terrorist
1,85 4,55 2,4 4,24 2,64 4,1 0 1 2 3 4 5 6 7
Negative Emotions Perceived Terrorist Threat
A = Terrorist Labelling C = Terrorist Labelling + Neutral Visual D = Terrorist Labelling + Threatening Visual
threat compared to the terrorist labelling (M = 4.55, SD = 1.21), nor did the threatening visual (M = 4.1, SD = 1.14) compared to the neutral visual or to the terrorist labelling. Hence, H3a and H3 b are rejected. Results are presented in Figure 2.
Anger and Fear
An analysis of variance (ANOVA) was used to test H4 whether the terrorist labelling affects fear stronger than anger a) compared to no labelling, b) when combined with a neutral visual and c) when combined with a threatening visual. Results showed that the conditions had a significant effect on anger, F(1,125) = 6.17, p = .014, ηp2 = .05. Simple contrasts
revealed that no labelling significantly increased anger (M = 2.58, SD = 2.12) compared to the terrorist labelling (M = 1.76, SD = 1.52). There was no significant difference between any condition on fear, F = 1.86.
Furthermore, simple contrasts revealed that the neutral visual significantly increased anger (M =2.49, SD = 1.85) compared to the terrorist labelling without visual (M = 1.76, SD
= 1.52), p = .027, 95 % CI [0.09, 1.38]. Compared to the neutral visual, the threatening visual
did not significantly increase anger (M = 2.65, SD = 2.12), but compared to the terrorist labelling without visual it did, p = .007, 95 % CI [0.24, 1.53]. Hypotheses H4a, H4b and H4c expecting that fear is induced stronger than anger are fully rejected.
The lack of increase of negative emotions in the terrorist labelling condition and the consequently comparable low levels of both negative emotions in the terrorist labelling conditions without visual (anger: M = 1.76, SD = 1.52; fear: M = 1.95, SD = 1.63) could potentially explain the significant difference between the terrorist labelling condition and the visual conditions on negative emotions and especially on anger. Hence, the visual conditions were compared to the condition without labelling for the purpose of further investigation. Then, simple contrasts show that there is no significant increase of anger in the neutral visual
condition (M = 2.49, SD = 1.85) nor in the threatening visual condition (M = 2.65, SD = 2.12) compared to the condition without terrorist labelling (M = 2.58). Results are presented in Figure 3.
Figure 3. The Effects of Newspaper Articles Varying in Labelling and Visuals on Fear and
Anger.
Note. Significance: * p < .05, **p < .01.
Discussion
This study sought to answer the question whether the terrorist labelling induces negative emotions as well as perceived terrorist threat and whether these effects increase when supplementing a neutral or a threatening visual within the methodological framework of an experimental online-study. Interestingly, the study’s theoretical expectations are refuted at several points, but findings enlighten the field of terrorism report studies.
Previous research on emotional responses to terrorism news has underlined the significance of the way terrorism is reported on as well as which information is provided. As such, it is well-known that news frames as well as emotionally laden information elicit emotions and increase perceived terrorist threat (Atwell Seate & Mastro, 2017; Kühne & Schemer, 2015; Lecheler, Schuck & de Vreese, 2013). Based on the relevance of linguistic choices (Obaidi, Kunst, Kteily, Sidanius & Thomsen, 2018; Slone, 2000), the present study focuses on news articles that refer to terrorism solely by terrorist labelling. By doing so, this study sheds light on an under-researched field of news effects. Theoretical considerations back up the expectation that terrorist labelling involves meta-narratives that evoke emotions and perceived threat (Shoshani & Slone, 2008). Findings reveal that the terrorist labelling did not have the expected positive influence on negative emotions and perceived terrorist threat. An obvious, although an alarming explanation is that German citizens are highly saturated by terrorism news and that habituation has occurred. Hence, it is possible that terrorist labelling or any other terrorism reference does not trigger negative emotions and perceived terrorist threat at all. Still, the results partially support the theoretical core assumption referring to terrorism’s semantical power given the fact that the level of negative emotions is higher for participants exposed to no labelling. As proven in prior research in similar contexts (Morin, 2016), it can be assumed that a news story dealing with an attack without labelling may still trigger terrorism associations and, consequently, positively impact negative emotions. Results of the manipulation check back up this guess: participants exposed to the article without labelling rather indicated that the story dealt with terrorism than vice versa. Nevertheless, the collected data does not allow for a strict empirical check of this assumption.
It is worth mentioning that participants possibly show stronger negative emotions and perceived terrorist threat in the condition without labelling especially due to the lack of information. They might have received the impression that some information was consciously
kept secret, which positively affected the appraisal of uncertainty and increased negative emotions in general. A broader contextualization of Western societies into the prevailing so-called Information Society based on Manuel Castells (2001) would underline this assumption. The Information Society is a response to the population's need for information and
knowledge, of which satisfaction should trigger the feeling of security, control and
consistency. Dissatisfaction of this need could explain the rise of negative emotions. Future research could manipulate the amount of information beyond terrorist labelling to broaden the scope of which news characteristics impact the reader’s emotions and perceptions of threat. Based on the dual-code-theory (Abraham & Appiah, 2006), the present study expects increasing effects on the dependent variables in multi-modal presentations using visuals. In the most general sense, this study confirms that the combination of terrorist labelling and visuals has the potential to increase effects on negative emotions. The findings have to be interpreted based on results of H1, namely that terrorist labelling did not significantly increase negative emotions and, hence, results are not fully in line with the theoretical expectation of a rise of effects. Still, as expected, the content of visuals significantly matters. In line with Roseman’s (1991) notion and Gadarian’s findings (2010) that the level of negativity inherent in events and especially in visuals affects the intensity of negative emotions, the terrorist labelling supplemented with a threatening visual significantly increased negative emotions compared to the article without visual. Even more appealing is that the pure presence of a visual (the neutral visual) matters to induce anger.
The most innovative finding of this study relates to the diverging effects of yielding anger and fear. In contrast to what was hypothesized, anger significantly increased as a response to news exposure, while fear did not. Anew, the news story without labelling significantly impacted anger. To explain this finding, the stimuli require examination. The message of all articles is somewhat positively in its core since the German police were able to
prevent a terrorist attack from happening. This might have transported the perception of control and certainty and clouded the uncontrollable aspects of terrorist attacks. In that case, anger is a consequence (Nabi, 1999; Turner, 2007). Moreover, prior studies have highlighted that anger is generally stronger in intensity than fear (Nabi, 2002; Newhagen, 1998), which is why it might emerge more clearly than fear in the current study.
This suggestion may lay the foundation for the consistent non-significant findings for perceived terrorist threat. Participants show higher levels of perceived terrorist threat than negative emotions throughout all conditions. Possibly, perceived terrorist threat can be considered a more stable and consistent perception similar to attitudes (Shoshani & Slone, 2008), which could explain lacking effects of exposure to only one news story. Future research should address this insight and assess perceived terrorist threat before and after experimental exposure. The scope of this study does not allow for a more in-depth evaluation of whether the relevance of the conceptualized determinants is overestimated. Earlier research has shown that manipulation of participant’s evaluation of events impacts their emotional responses (Iyer at al., 2014). Varying the assigned human agency as well as references to uncertainty and powerlessness in news articles in the future would offer a more concise understanding of relevant determinants. Given the small effect sizes of significant findings, the results require future replication. For the purpose of a comprehensive understanding, distinct positive visuals, as well as measurements of positive emotions could be subject to further investigation. Within a longitudinal survey-based study, Cho and colleagues (2003) have found that terrorism news elicits positive emotions on the long-term once emotions have enough time to unfold.
This study faces notable limitations. First, the self-reported measurement of emotions rises concerns about the strength of internal validity. Future studies may want to
of participant’s appraisals of the stimulus would have offered a better position to confirm or reject the relevance of Roseman’s determinants (Iyer et al., 2014; Powell, Boomgaarden, De Swert & de Vreese, 2015). Second, the sampling procedure restricts generalisation and is not representative of the German population, especially due to the use of SurveyCircle. Even though the choice of a snowball sampling strategy is justified by the fact that this study is interested in differing emotional reactions as a response to experimental exposure within an experimental methodology, the skewed distribution with regard to political ideology indicates that the sample represents a rather one-sided depiction of the German population. Future studies should attempt to recruit representative samples. Third, potential moderator variables are not taken into consideration. Aspects such as emotional attachment to the country of living, political attitudes and prior beliefs about terrorism can strongly influence emotions and perceived threat (Brouard, Vasilopoulos & Foucault, 2018; Conejero & Etxebarria, 2007). Fourth, the design of the study has to be considered. On the one hand, this study adds to the field of news effects in written and more crucially online news coverage. On the other hand, longer-lasting and more diverse exposure could reveal different findings. The news media tends to construct terrorist attacks as so-called media events (Nacos, Bloch-Elkon & Shapiro, 2007; Shoshani & Slone, 2008; Von Sikorski, Schmuck, Matthes & Binder, 2017). An extant exposure to terrorist labelling and differing imagery through several media sources would increase ecological validity and might yield stronger effects. Finally, general negative
emotions are assessed by summarizing anger and fear. This approach lacks completeness and does not cover a broad range of negative emotions.
To conclude, it is crucial to underline that this article does not aim at stipulating the news media how to report on events. It lies at the news media’s heart of democratic societies to provide truthful and detailed information. This study indicates that the role of the media is extensive, words and images are powerful components and news should report
well-considered and in a responsible manner. The results may help the news media as well as citizens and policymakers to understand how terrorist attacks unfold in societies by means of the media.
Endnotes
1Due to the most recent attacks, terrorism is highly associated with Islamic terrorism. It is crucial to note that there are many different forms of terrorism and that this study does not focus on a specific kind of ideological motivation.
2 It is important to note that, potentially, multiple emotions are evoked simultaneously (Nabi, 2002). Emotions are a complex concept and continuous subject of investigation mainly in psychology rather than in communication research. The conceptual definition provided in this article serves as a basic definition but does not claim to be fully comprehensive as this would extend the scope of this article. Moreover, there are studies indicating that correlations between positive and negative emotions occur, although they yield different effects (e.g., Brouard, Vasilopoulos & Foucault, 2018). 3 It is commonly known that news coverage tends to already include certain evaluations of
situations, which touches upon framing theory (Kühne & Schemer, 2015). Thus, individuals often do not evaluate situations solely based on personal processes. Mostly, they are exposed to content that is shaped by certain evaluations (Lecheler & de Vreese, 2018). For the sake of this study, the aspect of news framing is disregarded, and participants were exposed to digital written news that was not consciously set into a particular frame. Nevertheless, this aspect should be kept in mind since it corresponds to reality.
significant effects on emotions. Within his research, Roseman (1991) found that combinations of different determinants are possible, which potentially elicit different or stronger effects. The determinants human agency and uncertainty do not evoke considerably diverging effects when combined with other determinants. For a detailed elaboration, refer to Roseman, I. J. (1991). Appraisal determinants of discrete emotions. Cognition & Emotion, 5(3), 161-200.
5 From a theoretical viewpoint, it has to be noted that examples used to explain determinants of emotions in Appraisal Theory research almost exclusively refer to everyday situations (Watson & Spence, 2007; Roseman, 1991). As elaborated on, these determinants may especially hold for emotions toward terrorism, as it is such a meaningful and enormously momentous issue.
6 Recent public opinion data, as well as empirical findings, underpin this statement (Dürrholz, 2018; Johnson & Tversky, 1983; Zeit Online, 2017). Moreover, in line with terrorists’ strategy terrorist attacks can happen at any place and time, which yields a pessimistic perception of risk. Since terrorist attacks are aimed at symbolic values such as freedom and safety, individuals might feel attacked on the long-term. Sensational news and the production of attacks as media events contribute to this (Ortwin Renn, 2015 as cited in Kramer, 2015).
7 Prior research has shown considerable differences between anger and fear in terms of consequent behaviour. For instance, fear leads to a tendency of risk-averse behaviour and angry individuals rather use risk-seeking manners to solve problems (Dillard & Peck, 2001; Lerner & Keltner, 2001). While fear seems to decrease the willingness to process information, people tend to process information with increased attention and willingness when experiencing anger (Nabi, 1999). Since this study does not
investigate consequent behaviour, behavioural differences receive less attention in this context.
8 Given the fact that the study was conducted online, this technique is favourable for reaching a high number of participants within a brief period of time.
9 Next to practical considerations of studying online news articles, data justifies this decision. A study conducted by the Reuters Institute for the Study of Journalism (RI, 2018) investigating news consumption habits has found that the use of the internet as a news source is growing significantly in Germany (RI, 2018). Moreover, this study has proven unambiguously that most people prefer text when consuming online news, even showing a tendency towards a preference for less video material and more text news in Germany. The choice of the news outlet is justified by its popularity in the German news market. As likewise shown by RI (2018), Spiegel Online is the most frequently used online news outlet.
10 All measurements were conducted with an alpha-level of .05 and 95 % confidence intervals.
References
Abraham, L., & Appiah, O. (2006). Framing news stories: The role of visual imagery in priming racial stereotypes. Howard Journal of Communications, 17(3), 183–203. Armborst, A. (2010). Modelling terrorism and political violence. International
Relations, 24(4), 414-432.
Atwell Seate, A., & Mastro, D. (2017). Exposure to immigration in the news: The impact of group-level emotions on intergroup behaviour. Communication Research, 44(6), 817-840.
values in popular, quality, regional and financial newspapers. Journalism: Theory,
Practice & Criticism.
Brouard, S., Vasilopoulos, P., & Foucault, M. (2018). How terrorism affects political attitudes: France in the aftermath of the 2015–2016 attacks. West European
Politics, 2382, 1-27.
Castells, Manuel (2001): Das Informationszeitalter. Der Aufstieg der Netzwerkgesellschaft. Bd. 1, S. 407-459. Leverkusen: Leske und Buldrich Verlag.
Cho, J., Boyle, M. P., Keum, H., Shevy, M. D., McLeod, D. M., Shah, D. V., & Pan, Z. (2003). Media, terrorism, and emotionality: Emotional differences in media content and public reactions to the September 11th terrorist attacks. Journal of Broadcasting
& Electronic Media, 47(3), 309-327.
Cohrs, J. C., Kielmann, S., Maes, J., & Moschner, B. (2005). Effects of right‐wing
authoritarianism and threat from terrorism on restriction of civil liberties. Analyses of
Social Issues and Public Policy, 5(1), 263-276.
Conejero, S., & Etxebarria, I. (2007). The impact of the Madrid bombing on personal emotions, emotional atmosphere and emotional climate. Journal of social
issues, 63(2), 273-287.
Cooper, H. H. (2001). Terrorism: The problem of definition revisited. American Behavioral
Scientist, 44(6), 881-893.
Dillard, J. P., & Peck, E. (2001). Persuasion and the structure of affect. Dual systems and discrete emotions as complementary models. Human Communication Research, 27(1), 38-68.
Doosje, B., Zimmermann, A., Küpper, B., Zick, A., & Meertens, R. (2009). Terrorist threat and perceived Islamic support for terrorist attacks as predictors of personal and institutional out-group discrimination and support for anti-immigration policies–
Evidence from 9 European countries. Revue internationale de psychologie
sociale, 22(3), 203-233.
Dürrholz, J. (2018, May 7). Es kann auch mich treffen [Frankfurter Allgemeine Zeitung]. Retrieved from https://www.faz.net/aktuell/gesellschaft/menschen/angst-vor-terroristischen-anschlaegen-bei-jungen-menschen-15575826.html
Fahmy, S., Cho, S., Wanta, W., & Song, Y. (2006). Visual agenda-setting after 9/11: Individuals’ emotions, image recall and concern with terrorism. Visual
Communication Quarterly, 13(1), 4–15.
Gadarian, S. K. (2010). The politics of threat: How terrorism news shapes foreign policy attitudes. The Journal of Politics, 72(2), 469-483.
Geise, S., & Baden, C. (2014). Putting the image back into the frame: Modeling the
linkage between visual communication and frame-processing theory. Communication
Theory, 25(1), 46-69.
Global Terrorism Database (GTB) (2018). Data Collection Methodology. Retrieved from https://www.start.umd.edu/gtd/using-gtd/
Huddy, L., Feldman, S., Capelos, T., & Provost, C. (2002). The consequences of terrorism: Disentangling the effects of personal and national threat. Political Psychology, 23(3), 485-509.
Institut für politische Wissenschaft (2018). Größte Ängste der Deutschen nach Geschlecht im Jahr 2018. Retrieved from [Statista 2018]
https://de.statista.com/statistik/daten/studie/321259/umfrage/umfrage-zu-den-groessten-aengsten-der-deutschen/
Iyer, A., Webster, J., Hornsey, M. J., & Vanman, E. J. (2014). Understanding the power of the picture: The effect of image content on emotional and political responses to
Johnson, E. J., & Tversky, A. (1983). Affect, generalization, and the perception of risk. Journal of personality and social psychology, 45(1), 20.
Kramer, B. (2015). “Wir sollten risikomündiger werden” [fluter – Magazin der Bundeszentrale für politische Bildung]. Retrieved from
https://www.fluter.de/terrorangst-risikoforschung-wahrscheinlichkeiten
Kühne, R., & Schemer, C. (2015). The emotional effects of news frames on information processing and opinion formation. Communication Research, 42(3), 387–407.
Landau, M. J., Solomon, S., Greenberg, J., Cohen, F., Pyszczynski, T., Arndt, J… & Cook, A. (2004). Deliver us from evil: The effects of mortality salience and reminders of 9/11 on support for President George W. Bush. Personality and Social Psychology
Bulletin, 30(9), 1136-1150.
Lecheler, S., Schuck, A. R. T., & de Vreese, C. H. (2013). Dealing with feelings: Positive and negative discrete emotions as mediators of news framing effects. Communications -
The European Journal of Communication Research, 38(2), 189–209.
Lecheler, S., & de Vreese, C. H. (2018). News framing effects: Theory and practice. Routledge.
Lerner, J. S., & Keltner, D. (2001). Fear, Anger, and Risk. Journal of personality and social
psychology, 81(1), 146.
Lerner, J. S., Gonzalez, R. M., Small, D. A., & Fischhoff, B. (2003). Effects of fear and anger on perceived risks of terrorism: A national field experiment. Psychological
science, 14(2), 144-150.
Loewenstein, G. F., Weber, E. U., Hsee, C. K., & Welch, N. (2001). Risk as feelings. Psychological bulletin, 127(2), 267.
Marcus, G. E. (2000). Emotions in politics. Annual Review of Political Science, 3(1), 221- 250.
Morin, A. (2016). Framing terror: The strategies newspapers use to frame an act as terror or crime. Journalism & Mass Communication Quarterly, 93(4), 986–1005.
Nabi, R. L. (2002). Anger, fear, uncertainty, and attitudes: A test of the cognitive functional model. Communication Monographs, 69, 204-216.
Nabi, R. L. (1999). A cognitive‐functional model for the effects of discrete negative emotions on information processing, attitude change, and recall. Communication theory, 9(3), 292-320.
Nacos, B. L., Bloch-Elkon, Y., & Shapiro, R. Y. (2007). Post-9/11 terrorism threats, news coverage, and public perceptions in the United States. International Journal of
Conflict and Violence (IJCV), 1(2), 105-126.
Nellis, A. M., & Savage, J. (2012). Does watching the news affect fear of terrorism? The importance of media exposure on terrorism fear. Crime & Delinquency, 58(5), 748-768.
Newhagen, J. E. (1998). TV news images that induce anger, fear, and disgust: Effects on approach‐avoidance and memory. Journal of Broadcasting & Electronic Media, 42(2), 265-276.
Nerb, J., & Spada, H. (2001). Evaluation of environmental problems: A coherence model of cognition and emotion. Cognition & Emotion, 15(4), 521-551.
Obaidi, M., Kunst, J. R., Kteily, N., Sidanius, J., & Thomsen, L. (2018). Living under threat: Mutual threat perception drives anti-Muslim and anti-Western hostility in the age of terrorism. European Journal of Social Psychology, 48(5), 567-584.
O'neill, D., & Harcup, T. (2016). What is news?: News values revisited (again). Journalism
Studies, 18(12), 1470-1488.
Implications for right-wing attitudes and ethnic prejudice. International Journal of
Psychology, 48(1), 25-34.
Ottati, V. C., Steenbergen, M. R., & Riggle, E. (1992). The cognitive and affective components of political attitudes: Measuring the determinants of candidate evaluations. Political Behavior, 14(4), 423-442.
Powell, T. E., Boomgaarden, H. G., De Swert, K., & De Vreese, C. H. (2015). A clearer picture: The contribution of visuals and text to framing effects. Journal of
Communication, 65(6), 997–1017.
Powell, T. E., Boomgaarden, H. G., De Swert, K., & de Vreese, C. H. (2018). Framing fast and slow: a dual processing account of multimodal framing effects. Media
Psychology, 1–29.
Reuters Institute for the Study of Journalism (2018). Reuters Institute Digital News Report 2018. Retrieved from http://www.digitalnewsreport.org
Roseman, I. J. (1991). Appraisal determinants of discrete emotions. Cognition &
Emotion, 5(3), 161-200.
Schwarz, N., & Clore, G. L. (1996). Feelings and phenomenal experiences. Social
psychology: Handbook of basic principles, 2, 385-407.
Shoshani, A., & Slone, M. (2008). The drama of media coverage of terrorism: Emotional and attitudinal impact on the audience. Studies in Conflict & Terrorism, 31(7), 627–640. Slone, M. (2000). Responses to media coverage of terrorism. Journal of Conflict
Resolution, 44(4), 508-522.
Slovic, P. (1999). Trust, emotion, sex, politics, and science: Surveying the risk-assessment battlefield. Risk analysis, 19(4), 689-701.
Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive appraisal in emotion. Journal of
Stephan, W. G., & Stephan, C. W. (1996). Predicting prejudice. International Journal of
Intercultural Relations, 20(3-4), 409-426.
Stephan, C. W., & Stephan, W. S. (2013). An integrated threat theory of prejudice. In Reducing prejudice and discrimination (pp. 33-56). Psychology Press. Turner, M. M. (2007). Using emotion in risk communication: The anger activism
model. Public Relations Review, 33(2), 114-119.
Uenal, F. (2016). Disentangling islamophobia: The differential effects of symbolic, realistic, and terroristic threat perceptions as mediators between social dominance orientation and islamophobia. Journal of Social and Political Psychology, 4(1), 66-90.
Vasilopoulos, P., Marcus, G. E., & Foucault, M. (2018). Emotional responses to the Charlie Hebdo attacks: Addressing the authoritarianism puzzle. Political psychology, 39(3), 557-575.
Vasterman, P. L. (2005). Media-hype: Self-reinforcing news waves, journalistic standards and the construction of social problems. European Journal of Communication, 20(4), 508-530.
Von Sikorski, C., Schmuck, D., Matthes, J., & Binder, A. (2017). “Muslims are not
Terrorists”: Islamic State coverage, journalistic differentiation between terrorism and Islam, fear reactions, and attitudes toward Muslims. Mass Communication and
Society, 20(6), 825-848.
Watson, L., & Spence, M. T. (2007). Causes and consequences of emotions on consumer behaviour: A review and integrative cognitive appraisal theory. European Journal of
Marketing, 41(5/6), 487-511.
Yang, L., & Chen, H. (2018). Framing terrorist attacks: A multi-proximity model. International Communication Gazette, 81(5), 395–417.
from https://www.zeit.de/gesellschaft/zeitgeschehen/2017-09/studie-aengste-der-deutschen-2017-r-v-versicherung-umfrage
Appendix A Stimulus Materials
Condition C; Newspaper Article with Terrorist Labelling and Neutral Visual; Translated from German
Condition D; Newspaper Article with Terrorist Labelling and Threatening Visual; Translated from German
Appendix B Operationalization of Measurements
Demographics and personal characteristics
1. Age
What is your age (e.g. 30)? ______
2. Gender
Which gender describes you best? o Male
o Female
3. Education
What is the highest educational level that you have attained? [Adopted to German
educational system]
o Less than high school o High school graduate o Some college o 2-year degree o 4-year degree o Professional degree o Doctorate 4. Country of birth Where were you born?
o Germany
o Other, please specify ________________________________________________
5. Mother’s country of birth Where was your mother born?
o Germany
o Other, please specify ________________________________________________
6. Father’s country of birth Where was your father born?
o Germany
7. Prior victimization
Have you or a member of your household been the victim of a crime in the last 5 years (e.g. a burglary, assault)?
o Yes o No
8. News consumption
On an average day, how much time do you spend reading, watching or listening to the news on the following sources?
No time at all Less than 30 minutes 30 minutes - 1 hour More than 1 hour, up to 1,5 hours More than 1,5 hours, up to 2 hours More than 2 hours, up to 3 hours More than 3 hours Newspaper (printed) o o o o o o o TV news o o o o o o o Radio news o o o o o o o Digital news o o o o o o o Talk with friends, family or colleagues o o o o o o o
