The Influence of Geographical Location
and Type of News on Experienced Emotion
among Dutch Citizens
“An Emotional Window on the World”
Name: Leon Hollander Student number: 10562826 Mentor: Annemarie van Oosten Course: Bachelor Thesis
University: University of Amsterdam Word count: 7715
2
ABSTRACT
Previous research has brought to light that the current way foreign news is framed does not match with the interests of the public. This study examined this relation closer by doing an experiment in which the type of news and geographical location was manipulated (mixed methods design). No significant differences were found between substantive and sensational
news in Spain and Brazil. However, as both countries are highly similar it could be that the differences were masked and would surface when two completely different countries are being used. Another explanation might be that foreign news is always seen as foreign news independent of the distance to the home country. To research this the home country should be
added as a third variable. This current study does not support the belief that the way foreign news is portrayed in the media does not correspond with the interests of the public. Both type
of news regardless of the geographical location were equally effective in provoking affective responses. This has implications for the way foreign news should be portrayed by journalists, namely that they can decide themselves. The window through which we get to see the world has strong power and we are now one step closer in deciding which glass we are ought to use.
3
TABLE OF CONTENTS
Abstract ………2
Introduction ………..4
Theoretical Framework ………6
Type of News and Framing ………..6
Geographical Location and Framing ………7
Elaboration Likelihood Model ………..9
Method ……….10 Sampling ………...10 Procedure ………..10 Stimulus ………....11 Measures ………...12 Results ………..15 Discussion ………21 Literature ………..24 Research Instrument ……….28
4
“An Emotional Window on the World” INTRODUCTION
In 2015 there have been several terroristic attacks that were claimed by the Islamic State of Iraq and the Levant (ISIS). When we focus on two of them, the Paris attacks and the Beirut bombing, which both happened during the same week in November and had a similar amount of victims, there is a discrepancy in the amount of media attention both incidents received in the Dutch media. Journalists tend to write news about events that are more proximate to the home country (the process of domestication) (Clausen, 2004; Wilke, Heimprecht, & Cohen, 2012). At the same time, foreign news is increasingly framed and presented in a way that makes it more relevant to audiences in the home country (Clausen, 2004). It is the task of journalists to make the public aware of events that happen in the world. Journalists fulfil a gatekeeping function by deciding what is newsworthy for the domestic audience in the
endless flow of news from other sources of information (McNair, 2003). Journalists are by no means the only actors contributing to the construction of an image of the international reality. However, due the regularity, ubiquity and perseverance of news media, journalists should be seen as competitors for the number-one position in contributors in constructing an image of the international reality among citizens. Since actions are generally the result of an image and the beliefs someone hold (Galtung & Ruge, 1965), it is therefore of primary importance to research the image the news media give of the world.
Previous research by Lee, De Swert and Cohen (2015) concluded that the way foreign news is portrayed in the media does not correspond with the interests of the public. The research Lee, De Swert and Cohen (2015) conducted was a content analysis. They have established a difference in the framing of foreign news, but it is still unclear how those differences affect the public. As stated before, foreign news is increasingly framed and presented in a way that makes it in the eyes of journalists more relevant for the public to read. They do this by framing news in a sensational matter and relating the news-story with actors close to home (domestication) (Clausen, 2004). Sensational news outlets often make use of the different qualities of the human interest frame (Semetko & Valkenburg, 2000). This suggests that journalists hold the belief that the public do not care for news that is framed in a more substantive approach without involvement of any domestic actors. Substantive news outlets, in contrast predominantly use the different qualities of the responsibility frame and the conflict frame when framing news (Semetko & Valkenburg, 2000). This thesis will examine which type of news framing, sensational or substantive, is more effective in
5
provoking affective responses of the public regarding foreign news reporting. This contributes to science by adding to the scientific knowledge and understanding of the process of
domestication and the concept of foreign news.
Besides the type of news, the geographical location of the news article is manipulated. The sensational way in which news is framed does not match with the public demands
regarding foreign news, but does work for the framing of domestic news (De Swert & Cohen, 2013). One could expect that if the geographical location of the event depicted in the news article is located further away (so less domestic), a sensational coverage style is less effective. This possible relationship between geographical location of the news article, the type of news article and the message effectiveness has not been previously examined. It is important to study this relationship to establish if journalists should make a difference in the process of foreign news coverage. It could be that foreign news is always seen as foreign news by the public independent of the distance to the home country and one frame is the most effective. However, another possibility is that foreign news that is located geographically closer to the home country needs a different frame than foreign news that is located further away.
The effectiveness of a message can be measured cognitively or emotionally. In this thesis the effectiveness of the news articles are measured emotionally. This is for several reasons. When evaluating a stimulus cognitively the influence is found to be small (Chartrand & Bargh, 1999). Cognitive evaluations are characterized as the conscious evaluation of someone or something. Emotional and unconscious evaluations are found to be far more active and important in the process of evaluations than cognitive evaluations (Halberstadt & Niedenthal, 1997). In the past it has been the rule in political communication to study the cognitive media effects rather than affective media effects. Just recently there has been a switch in interest, researchers become more and more interested in studying emotional or affective response (Gross & Brewer, 2007; Nabi, 2003). This research will contribute to this relatively unexamined field of political communication by measuring the effectiveness of a message emotionally.
The following research question will be examined in this thesis: “What is the influence
of the geographical location and the type of news on the experienced emotion among Dutch citizens?”.
This thesis first examines the previous research that has been done within the concepts of framing, domestication and emotion and relevant theories will be discussed. After that, the method of this research will be described and explained thoroughly. This will be followed by
6
an analysis of the results which will lead to the conclusion of this research. Finally, the findings will be discussed and recommendations will be made for further research.
THEORETICAL FRAMEWORK An Introduction
In this section definitions of the different constructs are being discussed which then will be linked to previous research and theories that are relevant to the research question that is central in this thesis.
Framing
The process by which the media emphasize in salience of different aspects of a topic, is called framing (De Vreese, 2005). The way an issue is framed can influence one’s opinion and attitude (Lecheler, Schuck, & De Vreese, 2013). Chong and Druckman (2007)
distinguished two types of frames, equivalency and emphasis frames. In equivalency framing a news message is manipulated in such a way that you have a different presentation of the same issue. Researchers have found systematic differences in audience preference when the same problem was presented in different ways, such as gaining money or losing money (Quattrone & Tversky, 1988; Tversky & Kahneman, 1981). In emphasis framing different parts of an issue are presented to see which is more impactful. The audience can evaluate the issue described or the protagonists associated with the issues (Entman, 1993).
In this thesis the focus will lay on emphasis framing. Participants will be exposed to different aspects of the same issue. These aspects will vary on geographical location, the location of the event described will be either Brazil or Spain, and the type of news, a news article written in either a more substantive or sensational matter.
Type of news and framing.
One kind of framing is the type of news. Sensationalism in news is generally defined as news that is vivid and can capture one’s attention (Nisbett & Ross, 1980). News stories that are concrete and provoke an imagery are described as high in vividness (Nisbett & Ross, 1980). In a news story exemplary information is considered more concrete than general information (Zillmann, 2002). As a result, exemplars may be considered as sensational elements of news stories. Two types of exemplary information can be distinguished,
personalization and laypersons speaking. Reports about the personal situation of an individual may increase the concreteness of a news item. In news stories generally one can make a distinction between politicians and experts, and laypersons as comment maker. Comments
7
made by politicians and experts are considered mostly abstract and less concrete than comments made by laypersons (Hvitfelt, 1994). Comments made by politicians and experts can therefore be seen as less sensational and more substantive than comments made by laypersons. In addition, an active language style has been considered more concrete than a passive language style (Bostian, 1983). Previous research has found that sensationalist news outlets more often use the human interest frame (sensational), whereas sober and serious outlets more often use the responsibility and conflict frames (substantive) (Semetko & Valkenburg, 2000).
Geographical location and framing.
Another type of framing, next to type of news framing, is the geographical location. In the current study the geographical location of the protagonist -in which country did the event take place?- is being manipulated. Previous research on the effect of framing and variation of the geographical location has been limited to the context of immigration (Van Dijk, 1989; Cea D’Ancona, 2004; Van Gorp, 2005; Igartua et al., 2007). Most recently, Igartua,
Moral-Toranzo and Fernández (2011) examined the influence of news frames and group cues on how news about immigration is processed. They operationalized group cues as the protagonist of the information, which was someone with either Moroccan or Spanish heritage.
Participants either were exposed to a delinquency frame or an positive economic
consequences frame. One interaction effect was found between group cues and experienced negative emotions. The delinquency frame induced more negative emotions when the immigrants in question were said to be Morrocan than when they said to be Latin American. These findings contradict with findings made by Igartua and Cheng (2009), who found a stronger effect of group cues on cognitive, attitudinal and emotional variables. Igartua et al. (2011) give two possible explanations for these contradicting findings. The protagonist in the news article manipulation is only referred to incidental, no particular protagonist was
specified. The participants that took part in the research were highly interested in the topic of immigration (read: a high motivation) which lead to central processing of the news article. Whereas group cues have found to be cues that are processed peripheral (Brader et al., 2008; Cho et al., 2006; Domke, 2001).
Domestication.
The phenomenon in which foreign news is increasingly framed and presented in a way that makes it more relevant to audiences in the home country, is known as the domestication
8
of news (Clausen, 2004). Previous research on domestication has found that proximity is a strong determinant of the attention national media give to an event (Ruigrok & Van Atteveldt, 2007). Lee, De Swert and Cohen (2015) addressed the question of content-interest
correspondence (CIC) concerning purely foreign news, that is, the degree to which the types of content provided by the news media matches the content that the audience is interested in. They found that the CIC regarding the amount in which countries are being covered is high. In contrast, the CIC regarding the topics that are covered in foreign news are found to be low. The CIC regarding news topics are especially found to be low in countries where the media system leans towards commercialization and in which news outlets are more focused on soft news, this can be interpreted as a more sensational approach of media coverage. For domestic news the actual news coverage matches the interest of the public (De Swert & Cohen, 2013). One might expect that the coverage of foreign news is more impartial and less emotional when the home country is located further away from the location of the news event. However, previous research has found that this is not the case. Alasuutari, Qadir and Creutz (2013) examined the coverage of the Arab Spring in Britain, Finland and Pakistan and found that Britain and Finland wrote in a more sensational manner about the issue. Moreover, Lee, De Swert and Cohen (2015) found that concerning foreign news the types of content provided by the news media (which is sensationalized) does not match the content that the audience is interested in. Tai and Chang (2002) found a similar mismatch between audience interest and media coverage of events.
An explanation for these effects is the resonance theory (Gerbner et al., 1980). This theory argues that variables that are more approximate to someone are often conceptualized as direct experience. In turn, this conceptualization leads to a more intense reaction. When one is exposed with information that is located further away, previously held representations pay a smaller role which increases one’s chance to make an unbiased judgement. This might be the reason why one is more interested in a more substantive and objective news article about news in a location that is located further away from the home country. The resonance theory also argues that a frame that matches one’s interest is processed more elaborate (Gerbner et al., 1980). This theory has previously only been used in the context of television and violence. It is interesting to see if this theory can also be extended to other contexts.
Previous research has shown that news frames can result in different emotional reactions (Cho et al., 2003; Valentino et al., 2009).Since emotions depend on cognitive appraisals of facts or situations (Frijda, 1986) a more elaborate processing of information
9
should lead to a more intense emotional response as well. An emotion is a psychological state that consists of five components: cognitive appraisal or evaluation of a situation,
physiological arousal, motoric expression, motivation and a state of subjective feeling (Scherer, 1984). Given the effects of matching interests of media and audience concerning foreign news coverage, one could expect the following hypotheses:
Hypothesis I: When the geographical location of the news event is located further away from the home country a substantive style of news coverage will lead to a stronger emotional response, than a sensational style of news coverage.
Hypothesis II: When the geographical location of the news is located closer to the home country a sensational style of news coverage will lead to a stronger emotional response, than a substantive style of news coverage.
Elaboration likelihood model.
The way the news article is processed can affect (moderate) the effectiveness of the frames the news article consist of. The elaboration likelihood model suggests that there are two ways to process a message, central route processing and peripheral route processing (Petty & Cacioppo, 1986; Perse, 2001). When someone is exposed to a message, the level of elaboration determines through which route information is being processed. A high level of elaboration leads to central route processing of information, whereas a low level of
elaboration leads to peripheral route processing. A high level of elaboration is reached when one has the ability and motivation to process the message. Central route processing means that the receiver of the message tries to make a critical and rational driven evaluation of it to form arguments. Peripheral route processing, on the other hand, is superficial and automatic and based on peripheral cues. The route through which information is processed can affect the effectiveness of framing (Domke, 2001).
Group cues have found to be cues that are processed peripheral (Brader et al., 2008; Cho et al., 2006; Domke, 2001). This has implications for the current study. The main effect of geographical location (a peripheral cue) is expected to be stronger when information is processed through the peripheral route. The degree of interest in the topic of a news story is linked to greater cognitive elaboration (Petty & Cacioppo, 1986), the time a participant spend to analyze a news article is therefore an indicator to determine which processing route has been used. This has led to the following hypothesis.
10
Hypothesis III: The effect of geographical location on experienced emotion will be stronger for participants that processed the news article through the peripheral route, than for participants that processed the news article through the central route.
In relation to the elaboration likelihood model the use of a different type of news frame has several implications. When one processes the substantive and sensational news article through a central route and weighs the content, one should come to the conclusion that both issues are the same and just as good or bad. Sensational news could be experienced as more intense than substantive news when processed superficially due the more exaggerating writing style, which consists of many peripheral cues (e.g. adjectives) (Petty & Cacioppo, 1986). This has led to the following hypothesis.
Hypothesis IV: The effect of sensational news on experienced emotion will be stronger than the effect of substantive news for participants that processed the news article through the peripheral route.
METHOD Sampling
For recruiting participants convenience sampling was used (Patton, 2002). Participants were recruited on the social medium site Facebook or personally by phone. A total of 72 people filled out the survey completely (response rate: N/A; dropout rate: 20%). The mean age of the participants was 22.60 (SD = 6.76) with 33.3% being male. Convenience sampling has been criticized for its low external validity. However, Mullinix, Druckman and Freese (2014) have compared population-based samples with convenience samples and conclude that results are often highly similar and predictable in their dissimilarity. A huge benefit of
convenience sampling is the low amount of costs compared to population-based sampling. The only substantive criteria on which participants were selected was the variable age. Participants had to be at least eighteen years old. The participants are randomly assigned to each condition by Qualtrics. The researcher is a student himself and it was therefore expected that most of the participants would be students as well. The use of mostly students however does not create problems for causal inference of the found relationship between the
independent and dependent variables (Druckman & Kam, 2009).
Procedure
As this research is conducted to answer a cause-and-effect question, an experimental research strategy will be used (Gravetter & Forzano, 2012). By exposing one to a
11
manipulation but controlling for other variables you can conclude if the manipulation had an effect or not. The independent variables that have been manipulated in this study are the geographical location and the type of news. The stimulus material that has been manipulated is a news article. The independent variable geographical location was manipulated by
differing the location the story is set of an event. No connections were made to the Netherlands in the news article, as that could interfere with the independent variable geographical location due the process of domestication. The independent variable type of news (sensationalist versus substantive) was manipulated by using different frames and by using either a more active or passive writing style.
The subject of the news in the two types of news was kept the same. This was done to improve the internal validity of this research. It declines the possibility that the variation in the dependent variables are attributable to other causes than changes in the independent variables. Each independent variable consisted of two levels. The dependent variable emotion was measured implicitly as well as explicitly. As this experiment is interested in the effects on the dependent variable emotion and thus feelings, surveys have been found to be an effective method (Boeije, ‘t Hart & Hox, 2009). Questions were also asked about the control variables, need for affect and nationalism. The design that has been used is a mixed design. One will be exposed to both types of news (substantive versus sensational), but only to one geographical location (a European country versus a non-European country). By using a mixed design, less participants for the research were needed then when one had gone with a between subject design (Gravetter & Forzano, 2012). The within subject design element also decreased the role of individual differences. In addition, the between subject design decreased the role of time effects. To control for eventual time effects and order effects counterbalancing was used (Gravetter & Forzano, 2012).
Stimulus
The manipulation consisted of a news article. There were four different news articles (see appendix). The event depicted in the news articles took place in either a European country, Spain, or a non-European country, Brazil. Two different types of news articles were made for each country, a substantive and a sensational article. The substantive article has been framed in a responsibility and conflict frame. The sensational article has been framed in a human interest frame. Moreover, the language used in the substantive article is more passive than the language that has been used in the sensational article. To decide which frames to use a research led by Semetko and Valkenburg was used. In this research a content analysis was
12
done in which each frame was measured by establishing the presence of certain news
qualities. Those news qualities each reflected different frames. By manipulating a news article that satisfies on the strongest qualities of the human interest frame one could say that article matches a sensational news article. The strongest qualities of the human interest frame were found to be: providing a human face to the story, the use of adjectives or personal vignettes, emphasizing of how groups and individuals are affected by the issue and providing a story that goes into the personal life of actors. By manipulating a news article that satisfies on the strongest qualities of the responsibility and conflict frames one could say that article matches a substantive news article. The strongest qualities of the responsibility frame were found to be: suggesting in the story that the government could affiliate the problem, suggesting that the government is responsible for the problem and making suggestions in the story to solve the problem. The strongest qualities of the conflict frames were found to be: disagreement between parties/individuals/groups/countries, reproaching of one
party/individual/group/country to another and referring to two sides of the story.
Measures
Explicit emotions felt while reading the news article were measured with a 5-point Likert scale ( 1= not at all; 5 = very much). The scale consisted of ten items (Igartua, Cheng & Lopes, 2003) each measuring a different emotion. The ten discrete emotions that have been measured are disgust, contempt, anger, shame, fear, guilt, sadness, interest, surprise and happiness.
Implicit emotions felt while reading the news article were measured with a 4-point Likert scale (1= doesn’t match at all; 4 = matches perfectly) (Quirin, Kazén, & Kuhl, 2009). However here the emotional response was measured by asking participants how they thought neutral words (SUKOV and TALEP) matched with emotional adjectives (the IPANAT method). This is, in contrast to when one measures explicit emotions with a survey, a more unobtrusive measure and provides an extra dimension in the assessment of emotional intensity. The emotional adjectives that have been used in this study are happy, cheerful, energetic, helpless, tense and inhibited.
There is no indefinite operationalization for processing style as of yet. Processing style had been measured by the time someone was exposed to the stimulus as that is an indicator for greater cognitive elaboration (Petty & Cacioppo, 1986). To decide whether someone was processing either central or peripheral the median had been used as a measure of dispersion. A time slot was placed so participants were exposed to the news article for at least 15 seconds.
13
The median was measured and was taken as border for the two processing styles to create a new variable. Participants who spent less than 39.13 seconds observing the stimulus were assigned to the peripheral route category; participants who spend more time were assigned to the central route category.
Control Variables
Several variables were taken into account to control for as they could possibly distort the findings. In this research data was collected about the age, sex, nationalism and need for affect of participants. These variables are the control variables.
Need for affect.
Schuck and Otto (2014) found that the strength of the experienced emotion was moderated by the variable need for affect. Need for affect can be defined as the “general motivation of people to approach or avoid situations and activities that are emotion inducing for themselves or others” (Maio & Esses, 2001, p. 585). Need for affect of someone was measured with a 5-point Likert scale (1 = not at all; 5 = very much) based on the scale made by Appel, Gnambs and Maio (2012). The four items on which participants had to score themselves are: ‘It’s important for me to know what I feel.’, ‘I’m a very emotional person.’, ‘It’s important for me to know what other people feel.’ and ‘I think it’s important to reflect on my feelings.’. A factor analysis was done to compose a scale for need for affect. The four items that measured need for affect formed one component with an eigenwaarde of 2.23. All items had a factor load of more than 0.45. The component explains for 55.9% the variance in the four items. As the component is trustworthy a mean-scale could be made by averaging the items, α: .72, M = 3.58, SD = .73.
Nationalism.
Nationalism can be defined as a “notion that combines the political idea of territorial self-determination, the cultural idea of the nation as one’s primary identity and moral idea of justification of action to protect the rights of the nation against the other” (Ignatieff, 1993). Nationalism was measured with a single item scale in which one was directly asked how nationalistic they are (1 = not at all; 7 = very much).
Correlations.
Zero correlations were found between the variables nationalism or age and the
different explicit and implicit emotions. Someone’s sex is weakly correlated with the explicit emotion happiness, r = .19, p = .025, the explicit emotion anger, r = .18, p = .033 and the explicit emotion of fear, r = .20, p = .019. Females expressed stronger emotions of happiness anger and fear than males. Need for affect is strongly correlated with the explicit emotions
14
sadness, r = .41, p < .001, anger, r = .32, p < .001, and fear, r = .32, p < .001. Need for affect was moderately correlated with the explicit emotions disgust, r = .25, p = .003 and contempt,
r = .23, p = .007. Participants that scored higher on need for affect expressed stronger
emotions of sadness, anger, fear, disgust and contempt. The variables sex and need for affect are being controlled for during testing of the hypotheses.
Factor Analysis
Factor analysis were done to compose scales for the dependent variables. The seven items that measured negative explicit emotion formed two components. The first component has an eigenwaarde of 3.41 and explained for 48.7% the variance within its items. The second component has an eigenwaarde of 1.32 and explained for 18,9% the variance within its items. Component one consists of the items that measured the emotions sadness, anger, fear, disgust and contempt. The component is reliable and a scale was thus formed, α: .82, M = 2.51, SD = .90.. This scale can be interpreted as ‘negative explicit emotions(1)’. Component two consists of the items that measured the emotions guilt and shame. The component is reliable and a scale was thus formed, α: .83, M = 1.47, SD = .71.. This scale can be interpreted as ‘negative explicit emotions(2)’. The three items that measured positive explicit emotion formed one component. This component consists of the items that measured the emotions happiness, interest and surprise. The component has an eigenwaarde of 1.47 and explained for 48.8% the variance within its items. The initial Cronbach’s alpha was .44, but would increase if the item that measured happiness would be deleted. The scale for explicit emotions was then formed, α: .60, M = 2.89, SD = .93. This scale can be interpreted as ‘positive explicit emotions’. The emotion happiness will be analyzed separately. The six items that measured the implicit emotions one felt formed two components. The first component has an eigenwaarde of 3.03 and explained for 50.5% the variance within its items. The second component has an eigenwaarde of 1.24 and explained for 20.6% the variance within its items. Component one consists of the items that measured the characteristics happy, cheerful and energetic. The component is reliable and a scale was thus formed, α: .86, M = 2,04 SD = .76. This scale can be interpreted as ‘positive implicit emotions’. Component two consists of the items that measured the characteristics helpless, tense, inhibited. The initial reliability of this component was .69. However, when the item helpless is deleted Cronbach’s α increases to .80. A scale was formed, M = 2.40, SD = .86. This scale can be interpreted as ‘negative implicit
15
RESULTS
Before analyses were done, the data were cleaned. Participants that did not fill in the complete survey were deleted from the data. As there was made use of counterbalancing, one could presume that order effects disappear. Moreover, the measurement of the dependent variables were done completely independent of each other. To analyze the hypotheses of this thesis a new data file was therefore created in which the within component was split in separate cases. Each participant was represented two times in this new data file. This new file consisted of 144 cases and one could speak of a fully between subjects design now. The variable for geographical location (‘GL’) was manually constructed. If one was exposed to sensational and substantive articles of Brazil a participant received a ‘1’ on this variable – opposed to a ‘0’ for exposure to articles that occurred in Spain- (N Spain: 68; N Brazil: 73). Another variable was manually added, this variable made clear if a participant was exposed first to the sensational (‘0’) or substantive article (‘1’) (N sensational: 72; N substantive: 69). A time variable (‘tijdsvariabele’) was constructed by taking the total time every participant took to analyze the separate articles (M = 44.94, SD = 22.45). To determine outliers for this variable the outlier labeling rule was tested (Hoaglin, Iglewicz & Turkey, 1986). Three cases had a value above 126.60, the upper limit, and were therefore excluded from the data analysis (N=141).
Manipulation check
To check if people were aware to which manipulation they got exposed to two items were included at the end of the survey. The first item asked which country is located in the continent Europe. Everyone (100%) was aware that Spain is located inside of Europe and Brazil outside of Europe. The second item asked if one found the first or second news article more sensational. To analyze this item an independent t-test was executed. This test
concluded that the manipulation was significant and therefore successful. The participants remembered correctly which news article was more sensational, t(70) = 6.00, p < .000, 95% CI[.39, .78]. As meant they remembered when they were shown the substantive article (M = 1.23, SD = .43) and the sensational article (M = 1.81, SD = .40).
Analysis
To test the hypotheses a multivariate analysis was done. The dependent variables were all measured on interval level (numerical). The amount of independent variables was over two and were measured on categorical levels or higher. All groups have a similar amount of respondents so equal variance can be assumed. As one’s sex and need for affect was
16
significantly correlated with some dependent variables, they were added as covariates in the model and will be discussed now shortly before discussing the hypotheses.
When one focuses on the control variables, the following effects were found. One significant main effect was found between the independent variable sex and the dependent variables. One’s sex had a significant effect on the experienced emotion of happiness, F (1, 120) = 5.87, p = .017, n² = .04. Men (M = 1.26, SD = .61) experienced the emotion of happiness less than women (M = 1.57, SD = .86). Two significant main effect were found between the independent variable need for affect and the dependent variables. Need for affect had a significant effect on the experienced negative emotions: sadness, anger, fear, disgust and contempt, F (13, 120) = 2.26, p = .011, n² = .19. The higher one scored on the scale for need for affect, the stronger one experienced those negative emotions. Need for affect also had a significant effect on the experienced positive emotions of surprise and interest, F (13, 120) = 2.52, p = .004, n² = .21. The higher one scored on the scale for need for affect, the stronger one experienced those positive emotions. Wilks’ lambda was found to be significant for need for affect, F (91, 719) = 1.58, p = .001, but not for sex, F (7, 114) = 1.55, p = .158. Implicating that only the found effects for need of affect on emotions are valid.
Now focusing on the variables that were manipulated, geographical location and type of news, the following main effects were found. No main effects were found between the independent variable geographical location and the dependent variables (see table 1), Wilk’s lambda: F (7, 114) = 2.17, p = .042. Two main effects were found between the independent variable type of news and the dependent variables, Wilks’ lambda: F (7, 114) = 1.46, p = .189. Type of news had a significant effect on the implicit emotions: happy, cheerful and energetic, F (1, 120) = 6.46, p = .012, n² = .04. Sensational news (M = 2.20, SD = .78) resulted in more intense positive emotions than substantive news (M = 1.87, SD = .72). Type of news also had a significant effect on the implicit emotion helpless, F (1, 120) = 5.87, p = .017, n² = .04. Sensational news (M = 1.86, SD = .84) exposure led less to a feeling of feeling helpless than substantive news (M = 2.22, SD = .94). As Wilks’ lambda for type of news was not significant however, the found effects should be therefore taken into account with caution. Table 1. The influence of geographical location on experienced emotions.
Mean Square
df df Error F p
17 negative emotions Explicit negative emotions (1) 1.48 1 120 3.08 .082 Explicit positive emotions .01 1 120 .01 .932 Implicit positive emotions .00 1 120 .01 .940 Implicit negative emotions .08 1 120 .11 .738 Helpless .16 1 120 .19 .661 Happiness .56 1 120 .92 .339
Table 2. The influence of type of news on experienced emotions.
Mean Square df df Error F p Explicit negative emotions .97 1 120 1.36 .247 Explicit negative emotions (1) .02 1 120 .05 .832 Explicit positive emotions .62 1 120 .81 .371 Implicit negative emotions 1.80 1 120 2.46 .120 Happiness .20 1 120 .33 .567
18
Now the hypotheses were tested. The first hypothesis stated that “When the
geographical location of the news event is located further away from the home country a substantive style of news coverage will lead to a stronger emotional response, than a sensational style of news coverage.”. The second hypothesis stated that “When the
geographical location of the news is located closer to the home country a sensational style of news coverage will lead to a stronger emotional response, than a substantive style of news coverage.” To test those hypotheses the interaction between the variables type of news and
geographical location on the dependent variables was measured. No significant interaction effects were found (see table 3), Wilks’ lambda: F (7, 114) = 1.13, p = .349. With that, both hypothesis are rejected.
Table 3. The influence of the interaction between geographical location and type of news on
experienced emotions. Mean Square df df Error F p Explicit negative emotions .48 1 120 .68 .413 Explicit negative emotions (1) .02 1 120 .04 .848 Explicit positive emotions 2.74 1 120 3.54 .062 Implicit positive emotions .82 1 120 1.46 .228 Implicit negative emotions .37 1 120 .512 .476 Helpless .09 1 120 .11 .739 Happiness .53 1 120 .87 .352
19
To test the third hypothesis the interaction between geographical location and
information processing route on the dependent variables was measured. The third hypothesis stated that “The effect of geographical location on experienced emotion will be stronger for
participants that processed the news article through the peripheral route, than for participants that processed the news article through the central route.”. No significant
interaction effect was found (see table 4), Wilks’ lamda: F (7, 114) = 0.86, p = .543.
Hypothesis III was therefore rejected. Moving to the final hypothesis, the interaction between type of news and information processing route on the dependent variables was measured. The fourth hypothesis stated that “The effect of sensational news on experienced emotion will be
stronger than the effect of substantive news for participants that processed the news article through the peripheral route.”. One significant interaction effect was found. Type of news
and information processing route had a significant interaction effect on the implicit emotions of feeling tense and inhibited, F (1, 120) = 5.29, p = .023, n² = .04. Participants that
processed news articles through the central route experienced the implicit emotions of feeling tense and inhibited somewhat higher when exposed to a sensational news article (M = 2.53,
SD = .81) than when exposed to a substantive news article (M = 2.43, SD = .83). For
participants that processed news articles through the peripheral route the experienced emotion of feeling tense and inhibited was less intense when exposed to a sensational news article (M
= 2.09, SD = .88) compared to when one got exposed to a substantive news article (M = 2.63, SD = .88). This is opposite to what was expected and thus hypothesis IV was rejected
(concerning negative implicit emotions). No other significant interaction effects were found (see table 6). Hypothesis IV has been rejected completely. Wilks’ lambda was found to be not significant, F (7, 114) = 1.23, p = .293, the found effect should therefore be taken into
account with caution.
Table 4. The influence of the interaction between geographical location and information
processing route on experienced emotions. Mean Square df df Error F p Explicit negative emotions .30 1 120 .42 .517
20 Explicit negative emotions (1) .00 1 120 .00 .954 Explicit positive emotions .57 1 120 .74 .393 Implicit positive emotions .02 1 120 .03 .869 Implicit negative emotions 1.08 1 120 1.48 .226 Helpless 1.84 1 120 2.26 .135 Happiness .17 1 120 .28 .597
Table 5. The influence of the interaction between type of news and information processing
route on experienced emotions. Mean Square df df Error F p Explicit negative emotions .01 1 120 .01 .922 Explicit negative emotions (1) .04 1 120 .09 .765 Explicit positive emotions .16 1 120 .20 .652 Implicit positive emotions .00 1 120 .00 .973 Helpless .06 1 120 .07 .787
21
Happiness 1.17 1 120 1.92 .169
DISCUSSION
The question asked at the beginning of this thesis was: “What is the influence of the
geographical location and the type of news on the experienced emotion among Dutch citizens?”. To answer this question four hypotheses were formulated. There was found no
significant interaction between geographical location and type of news on the emotions one experienced. This finding contradicts with the resonance theory. Research in which the resonance theory was previously used was always limited to the context of television and violence. This research was conducted in the context of online news with as common theme a fire in a hotel. It seems that the resonance theory cannot be extended to other contexts than television and violence for now. Nor was there found a significant interaction between geographical location and the way information was processed on the emotions one
experienced. The reason why no interaction effects were found with processing style could be due the way processing style was operationalized. In further research it might be better to add a question in which participants are asked at the end of the survey how concentrated they were. Although measuring processing style would still be rather arbitrary. The effect of sensational news on experienced emotion was found not to be stronger than the effect of substantive news for participants that processed the news article through the peripheral route. No main effects were found for the geographical location on experienced emotion. Some main effects were found for the type of news on experienced emotion. Positive emotions were felt more intense when reading a sensational news article, whereas negative emotions were felt more intense when reading a substantive news article.
These findings contribute to science for several reasons. Lee, De Swert and Cohen (2015) established a difference in the framing of foreign news. This research expands on those findings by making clear how those different frames affect the public emotionally. Also the effectiveness of a message was measured emotionally instead of cognitively in this research. This is a relatively unexamined field of political communication. Expected was that if the geographical location of the event depicted in the news article is located further away a sensational coverage style would be less effective than substantive coverage style. However, the findings made in this research do not support that assumption. No significant interaction and therefore difference was found in the relationship between geographical location and type of news on the experienced emotion. A possible explanation why no differences were found is
22
due the choice of the countries when manipulating the geographical location. Brazil and Spain are cultural identical and although the geographical location differs –and people were aware of this difference in distance- the associations made in the mind when one is exposed to those geographical locations might be too similar as opposed to when one had got exposed to for example Switzerland and Kenya which are different on much more aspects.
Another explanation why no differences were found might be due the
operationalization of the concept domestication. In this research the assumption was that foreign news that is closer to home is more domestic than news that is further from home. However, it could be that foreign news - no matter how close to or far from the home country is seen as the same, which is foreign news. If that is the case, the findings made in this
research would theoretically make more sense (possibly supporting all the hypothesis). Furthermore, it would provide a starting point for future research. In this research one would add a third geographical location, namely the home country. The formulated hypotheses of this thesis would be altered in such way that the expected effects for news closer to home only applies for news that happens in the home country itself. When one compares the means for each emotion separately in relation to the different conditions of the independent variables, one can see that there is no clear pattern and the differences in experienced emotions are rather small. The biggest differences in the experienced emotion are seen when one was exposed to a different type of news coverage style. This suggests again that the manipulation of the geographical location did not spark different responses, whereas the manipulation of the type of news did. When comparing the means for each emotion in relation to the
interaction of the independent variables, again no evident pattern was observed. There was no indication that the hypotheses were formulated in the right direction, but just failed to be significant.
There are several methodological limitations for this research. Participants were exposed to two news articles each, this could have led to time effects and systematic bias. To control for those effects counterbalancing was used. The control over the environment where the experiments took place in this study was low. The survey was spread online, participants could have done the survey in any location. This increases the external validity of the
experiment, however decreases the internal validity. Participants might not have been completely focused when doing the survey which has led to distortion of the results. Other variables in the environment and the device on which the survey was filled in could have played as a confounding variable in this study as well. For future research I would suggest a more artificial setting. To be sure that participants are concentrated when doing the survey a
23
control question that should be left blank could be added.
Previous research has brought to light that the current way foreign news is framed does not match with the interests of the public. This study examined this relation closer by doing an experiment in which the type of news and geographical location was manipulated. No significant differences were found between substantive and sensational news in Spain and Brazil. This study does not support the belief that the way foreign news is portrayed in the media does not correspond with the interests of the public. Both type of news regardless of the geographical location were equally effective in provoking affective responses. This has
implications for the way foreign news should be portrayed by journalists, namely that they can decide themselves. The window through which we get to see the world has strong power and we are now one step closer in deciding which glass we are ought to use.
LITERATURE
Alasuutari, P., Qadir, A., & Creutz, K. (2013). The domestication of foreign news : news stories related to the 2011 Egyptian revolution in British , Finnish and Pakistani newspapers. Media, Culture & Society, 35(6), 692–707.
24
Appel, M., Gnambs, T., & Maio, G. (2012). A short measure of the need for affect. Journal of
Personality Assessment, 94(4), 418-426.
Bostian, L. R. (1983). How active, passive and nominal styles affect readability of science writing.Journalism and Mass Communication Quarterly,60(4), 635.
Brader, T. (2005). Striking a responsive chord: How political ads motivate and persuade voters by appealing to emotions. American Journal of Political Science, 49, 388–405. Brader, T., Valentino, N. A., & Suhay, E. (2008). What triggers public opposition to
immigration? Anxiety, group cues, and immigration threat.American Journal of
Political Science, 52(4), 959-978.
Chartrand, T. L., & Bargh, J. A. (1999). The chameleon effect: The perception–behavior link and social interaction. Journal of personality and social psychology, 76(6), 893. Cho, J., Boyle, M., Keum, H., Shevy, M., McLeod, D., Shah, D., & 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.
Cho, J., De Zuniga, H. G., Shah, D. V., & McLeod, D. M. (2006). Cue Convergence
Associative Effects on Social Intolerance. Communication Research,33(3), 136-154. Chong, D., & Druckman, J. (2007). Framing Theory. Annual Review of Political Science, Clausen, L. (2004). Localizing the Global: ‘Domestication’ Processes in International News
Production. Media, Culture & Society, 26(1), 25–44.
Cohen, A. A. (Ed.). (2013). Foreign News on Television: Where in the World is the Global
Village?.
De Vreese, C. H. (2005). News framing: Theory and typology. Information design journal+
document design, 13(1), 51-62.
Domke, D. (2001). Racial Cues and Political Ideology An Examination of Associative Priming.Communication Research, 28(6), 772-801.
Druckman, J. N., & Kam, C. D. (2009). Students as experimental participants: A defense of the 'narrow data base'. Available at SSRN 1498843.
25
Entman, R. M. (1993). Framing: Toward clarification of a fractured paradigm. Journal of
communication,43(4), 51-58.
Galtung, J., & Ruge, M. H. (1965). The structure of foreign news the presentation of the Congo, Cuba and Cyprus Crises in four Norwegian newspapers. Journal of peace
research, 2(1), 64-90.
Gerbner, G., Gross, L., Morgan, M., & Signorielli, N. (1980). The “mainstreaming” of America: Violence profile no. 11. Journal of communication, 30(3), 10-29. Gravetter, F. J., & Forzano, L. B. (2012). Research methods for the behavioral sciences.
Wadsworth, Cengage Learning. International Edition: ISBN-13, 978-1. Gross, K., & Brewer, P. R. (2007). Sore losers: News frames, policy debates, and
emotions. The Harvard International Journal of Press/Politics, 12(1), 122-133. Halberstadt, J. B., & Niedenthal, P. M. (1997). Emotional state and the use of stimulus
dimensions in judgment. Journal of Personality and Social Psychology, 72(5), 1017.
Hoaglin, D.C., Iglewicz, B., and Tukey, J.W. (1986). Performance of some resistant rules for outlier labeling, Journal of American Statistical Association, 81, 991-999.
Hvitfelt, H. (1994). The commercialization of the evening news: Changes in narrative technique in Swedish TV news. Nordicom review, 15(1), 33-41.
Igartua, J. J., Cheng, L., & Lopes, O. (2003). To think or not to think: Two pathways towards persuasion by short films on AIDS prevention. Journal of Health
communication, 8(6), 513-528.
Igartua, J. J., Moral-Toranzo, F., & Fernández, I. (2012). Cognitive, attitudinal, and emotional effects of news frame and group cues, on processing news about immigration. Journal
of Media Psychology.
Ignatieff, M. (1993). The Balkan Tragedy. The New York Review of Books, 40(9), 13. Lecheler, S., Schuck, A. R., & de Vreese, C. H. (2013). Dealing with feelings: Positive and
26
Lee, F., De Swert, K., & Cohen, A. A. (2015). Do the contents of foreign news on television match viewers’ interests? A 12-nation study of topics and countries of interest.
Communication Research.
Maio, G., & Esses, V. (2001). The Need for Affect: Individual differences in the motivation to approach or avoid emotions. Journal of Personality, 69(4), 584-615.
Mullinix, K. J., Leeper, T. J., Druckman, J. N., & Freese, J. (2015). The generalizability of survey experiments. Journal of Experimental Political Science, 2(02), 109-138. McNair, B. (2011). An introduction to political communication. Taylor & Francis. Nabi, R. L. (2003). Exploring the framing effects of emotion do discrete emotions
differentially influence information accessibility, information seeking, and policy preference?. Communication Research, 30(2), 224-247.
Nisbett, R. E., & Ross, L. (1980). Human inference: Strategies and shortcomings of social judgment.
Patton, M. Q. (2002). Designing qualitative studies. Qualitative research and evaluation
methods, 3, 230-246.
Perse, E. M. (2001). Media effects and society. Routledge.
Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion.
Advances in Experimental Social Psychology, 19, 124-192.
Quattrone, G. A., & Tversky, A. (1988). Contrasting rational and psychological analyses of political choice. American Political Science Review, 82(03), 719-736.
Quirin, M., Kazén, M., & Kuhl, J. (2009). When nonsense sounds happy or helpless: The Implicit Positive and Negative Affect Test (IPANAT). Journal of personality and
social psychology, 97(3), 500.
Ruigrok, N., & Van Atteveldt, W. (2007). Global angling with a local angle: How US, British, and Dutch newspapers frame global and local terrorist attacks. The Harvard
International Journal of Press/Politics, 12(1), 68-90.
Scherer, K. R. (1984). Emotion as a multicomponent process: A model and some cross cultural data. Review of Personality & Social Psychology.
27
Schuck, A.R.T., & Otto, L.P. (2014). Framing climate change - emotional responses as
mediators for the effect of threat framing on individual risk perceptions and behavioral intentions. Paper presented at the annual conference of the International
Communication Association (ICA), Seattle/USA.
Semetko, H. A., & Valkenburg, P. M. (2000). Framing European politics: A content analysis of press and television news. Journal of communication, 50(2), 93-109.
Shrum, L. J., & Bischak, V. D. (2001). Mainstreaming, resonance, and impersonal impact.
Human Communication Research, 27(2), 187-215.
Tai, Z., & Chang, T. K. (2002). The Global News and the Pictures in Their Heads A Comparative Analysis of Audience Interest, Editor Perceptions and Newspaper Coverage. International Communication Gazette, 64(3), 251-265.
Tversky, A., & Kahneman, D. (1981). The framing of decisions and the psychology of choice. Science,211(4481), 453-458.
Valentino, N.A., Gregorowicz, K., & Groenendijk, E.W. (2009). Efficacy, emotions and the habit of participation. Political Behavior, 31, 307-330.
Van Dijk, N. M. (1989). The importance of bias-terms for error bounds and comparison results.
Van Gorp, B. (2005). Where is the frame? Victims and intruders in the Belgian press coverage of the asylum issue. European Journal of Communication,20(4), 484-507. Wilke, J., Heimprecht, C., & Cohen, A. A. (2012). The geography of foreign news on
television: A comparative study of 17 countries. International Communication
Gazette, 74, 301-322.
Zillmann, D. (2002). Exemplification theory of media influence. Media effects: Advances in
theory and research, 2, 19-41.
28
Manipulations
Substantive articles:
“
ONDUIDELIJKHEID OVER OORZAAK EXPLOSIE SPANJE”
ZARAGOZA, Spanje – Het is nu enkele weken na de tragedie in Spanje. Het aantal
slachtoffers na de explosie in het Spaanse 5-sterren hotel ‘Corona de Arágon’ is
inmiddels opgelopen tot 80 mensen. Nog nooit heeft een brand zoveel levens opgeëist in
Spanje. De Spaanse overheid heeft laten weten geschokt te zijn over het voorval. Het is
vooralsnog onbekend of de explosie incidenteel van aard is, of dat de terreurorganisatie
ETA hier achter zit.
In een officiële verklaring van de Spaanse overheid laat zij weten dat de brand
incidenteel van aard is en gestart is door een gasexplosie in het café van het hotel.
Echter, uit gelekte documenten van onderzoek naar het incident blijkt dat er sporen van
de stof Napalm zijn gevonden. Deze stof wordt voornamelijk gebruikt in brandbommen
en is gevreesd voor de ernstige brandwonden die het bij mensen achterlaat. De Spaanse
overheid heeft hier nog niet op gereageerd.
“
ONDUIDELIJKHEID OVER OORZAAK EXPLOSIE BRAZILIË”
SÃO PAULO, Brazilië – Het is nu enkele weken na de tragedie in Brazilië. Het aantal
slachtoffers na de explosie in het Braziliaanse 5-sterren hotel ‘Corona de Arágon’ is
inmiddels opgelopen tot 80 mensen. Nog nooit heeft een brand zoveel levens opgeëist in
Brazilië. De Braziliaanse overheid heeft laten weten geschokt te zijn over het voorval.
Het is vooralsnog onbekend of de explosie incidenteel van aard is, of dat de
terreurorganisatie FARC hier achter zit.
In een officiële verklaring van de Braziliaanse overheid laat zij weten dat de brand
incidenteel van aard is en gestart is door een gasexplosie in het café van het hotel.
Echter, uit gelekte documenten van onderzoek naar het incident blijkt dat er sporen van
de stof Napalm zijn gevonden. Deze stof wordt voornamelijk gebruikt in brandbommen
29
en is gevreesd voor de ernstige brandwonden die het bij mensen achterlaat. De
Braziliaanse overheid heeft hier nog niet op gereageerd.
Sensational articles:
“
IK HOOPTE DAT HET EEN NACHTMERRIE WAS, MAAR IK WERD NIET
WAKKER”
ZARAGOZA, Spanje – Meer dan 300 gasten verbleven in het Spaanse 5-sterren hotel
‘Corona de Arágon’ tijdens de nacht van de brand. De gruwelijke brand brak uit na een
explosie in de keuken van het naburige café. Autoriteiten hebben laten weten dat er
minstens 80 mensen om het leven zijn gekomen tijdens deze tragedie.
De Spaanse weduwe Carmen Polo de Franco (31), verbleef tijdens de ramp op de
tweede verdieping van het hotel en heeft het overleefd. ‘Ik lag te slapen toen ik plotseling
een harde explosie hoorde gevolgd door een heleboel geschreeuw. Ik liep naar de gang
en alles was rood, rood van het vuur en rood van bloed. Een man die mij in de ochtend
vriendelijk begroette, lag nu verminkt op de grond. Ik werd gek. Ik hoopte dat het een
nachtmerrie was, maar ik werd niet wakker.’ Carmen wist te ontsnappen via de
nooduitgang, waar zij samen met andere lotgenoten werd opgevangen door de
ambulance.
“
IK HOOPTE DAT HET EEN NACHTMERRIE WAS, MAAR IK WERD NIET
WAKKER”
SÃO PAULO, Brazilië – Meer dan 300 gasten verbleven in het Braziliaanse 5-sterren
hotel ‘Corona de Arágon’ tijdens de nacht van de brand. De gruwelijke brand brak uit na
een explosie in de keuken van het naburige café. Autoriteiten hebben laten weten dat er
minstens 80 mensen om het leven zijn gekomen tijdens deze tragedie.
De Braziliaanse weduwe Carmen Polo de Franco (31), verbleef tijdens de ramp op de
tweede verdieping van het hotel en heeft het overleefd. ‘Ik lag te slapen toen ik plotseling
een harde explosie hoorde gevolgd door een heleboel geschreeuw. Ik liep naar de gang
en alles was rood, rood van het vuur en rood van bloed. Een man die mij in de ochtend
30
vriendelijk begroette, lag nu verminkt op de grond. Ik werd gek. Ik hoopte dat het een
nachtmerrie was, maar ik werd niet wakker.’ Carmen wist te ontsnappen via de
nooduitgang, waar zij samen met andere lotgenoten werd opgevangen door de
ambulance.
Wat is uw geslacht? Wat is uw leeftijd? Explicit emotions:
Zou u nu kunnen aangeven in welke mate u de volgende emoties nu ervaart? (1 = totaal niet; 5 = heel erg) Blijdschap Verdriet Woede Angst Walging Minachting Schuld Schaamte Interesse Verbazing Implicit emotions:
In het volgende onderdeel zijn wij benieuwd hoe de klanken van bepaalde niet bestaande woorden worden geïnterpreteerd.
Geef nu aan in hoeverre u de volgende eigenschappen bij het woord ‘SUKOV’ vindt passen. 1 = past er totaal niet bij, 2 = past er een beetje bij, 3 = past er redelijk goed bij 4 = past er goed bij
Blij Vrolijk Energiek Hulpeloos
31
Gespannen Onderdrukt
Geef nu aan in hoeverre u de volgende eigenschappen bij het woord ‘TALEP’ vind passen. 1 = past er totaal niet bij, 2 = past er een beetje bij, 3 = past er redelijk goed bij 4 = past er goed bij Blij Vrolijk Energiek Hulpeloos Gespannen Onderdrukt Need for affect:
Zou u nu kunnen aangeven in welke mate de volgende stellingen in het algemeen overeenkomen met uzelf? (1 = totaal niet; 5 = heel erg)
Het is belangrijk voor mij om te weten wat ik voel. Ik ben een heel emotioneel persoon
Het is belangrijk voor mij om te weten hoe anderen zich voelen. Ik vind het belangrijk om te reflecteren op mijn gevoelens. Nationalism:
Hoe nationalistisch ziet u zichzelf? (1 = totaal niet nationalistisch; 7 = zeer sterk nationalistisch)
Manipulationcheck: Welk land ligt in Europa? Spanje
Brazilië
Welke artikel vond u meer sensationeel geschreven? Ik vond het eerste artikel meer sensationeel geschreven. Ik vond het tweede artikel meer sensationeel geschreven.
