Audience Motivations for
Reading Sensationalized
News Articles
Master’s thesis by Lola Lindhout
Student Number: 10092498 University of Amsterdam
Graduate School of Communication
Master’s Programme Communication Science: Political Communication and Journalism Supervised by Jonas Lefevere
Abstract
As scholars debate the possible impact of sensationalized political news articles on citizens’ democratic functioning, news outlets feel growing pressures to produce political news in this format. Market pressures compel them to answer to consumer demand, which often is sensationalistic. This study aims to investigate the audience’s motivations to prefer sensationalized political news articles over non-sensationalized political news articles. The audience motivations are grouped into four typologies: informational guidance, surveillance, entertainment and social utility. A web-based survey was conducted among 85 Dutch citizens of the ages 21 and older. Logistic regression shows that informational guidance motivations lead people to prefer sensationalized news articles, but surveillance motivations lead people to prefer the opposite. Results further show that the role of social utility motivations is still uncertain, and should be further investigated.
Introduction
“All through the hours of going to press that night, the Harmsworth watchword was: ‘Keep it down gentlemen!’ He moved constantly through the newsroom repeating his injunction to be brief: ‘No story longer than two hundred and fifty words!’ ” (Pound & Harmsworth, 1959: 266)
Alfred Harmsworth (later known as Lord Northcliffe) was the first journalist to use the word “tabloid” in a journalistic sense. He borrowed the word from a manufacturer of pills (combining the terms tablet and alkaloid), as it resembled his vision for his new newspaper, the Daily Mail (est. 1896): small, concentrated and effective. He referred to his newspaper as ‘the daily time saver’, and argued this news format fitted the quick-paced twentieth century (Örnebring & Jönsson, 2007). Tabloid journalism has since been growing in the Western world and is seen as an effect of the broader trend of news commercialization (McManus, 2009). Due to this commercialization, news outlets feel increased pressure to respond to consumer demand. The shift towards a more consumer-oriented style of news production has been associated with the reoccurring notice that news articles have been showing more signs of tabloidization over time (Blumler & Gurevitch, 1995; Blumer & Kavanagh, 1999; McManus, 1994; Grabe et al., 2001; Uribe & Gunter, 2007). When looking at political news coverage in particular, these signs of tabloidization also appear. An important aspect of the tabloidization of political news is sensationalism. Sensationalism appeals to consumer demand, and it refers to a writing style that is evoking, emotional and sensory stimulating (Grabe et al., 2001).
This sensational writing style triggered a debate regarding its implications for democracy. Some say it threatens democracy because it breeds cynicism and reduces political
interest, while avoiding the actual political issues to focus more on political scandals (Örnebring & Jönsson, 2007). But on the other hand, there are scholars who debate the possibility that sensational journalism provides an ‘alternative’ public sphere, because it reaches another segment of the audience that might be less interested in politics. This alternative public sphere would engage the less politically interested public in politics, using a more accessible writing style (Örnebring, 2007; Örnebring & Jönsson, 2007). As it may have significant implications for the way citizens function in democratic societies, sensationalism is an important field of study (Davis & McLeod, 2003; Grabe, Zhou & Barnett, 2001; Hofstetter & Dozier, 1986; Semetko & Valkenburg, 2000; Sparks & Tulloch, 2000; Uribe & Gunter, 2007).
The relationship between the preferences of these more sensation-oriented consumers and journalistic content has been studied by several scholars (Hamilton, 2004; Esser, 1999; Franklin, 1996; McLachlan & Golding, 2000; Metykova, 2008). However, the reasons that drive people to prefer sensationalistic news have not been investigated. Previous studies have investigated audience motivations for preferring certain media outlets in the search for political information (Kaye & Johnson, 2002; Kaye & Johnson, 2004; Lin, 2006; Yang, 2004), but the motivations for preferring this political information in a certain article format have not been studied. However, different types of media outlets argue that they are confronted with the same sensationalistic audience demand (Metykova, 2008). Taking into consideration the possible impact answering to this sensationalistic demand has for the political engagement of citizens, this study attempts to reduce this scientific knowledge gap by investigating the audience’s motivations for preferring a sensational writing format. Based on a study by Hamilton (2004) and with insights of the Uses and Gratifications approach (McQuail, 1994; Ruggiero, 2000), a typology of motivations is sought. The practical implications for this study can be of interest for the journalistic field, as the results could
provide further insight in consumer demand. As such, the following research question is formulated:
“What are the audience’s motivations to prefer sensationalised political news articles over non-sensationalized political news articles?”
The thesis proceeds as follows: firstly, a theoretical overview discussing news commercialization, consumer demand, sensationalism and audience motivations is provided, followed by a methods section after which the results of the study will be discussed. The thesis ends with a discussion of the conclusions, implications and limitations of the present study.
Theoretical Framework
News commercialization
News organizations have always been dependent on selling (McManus, 2009). However, claims of commercialization of the news media, which imply a shift in a news organization’s focus from public-sense making towards maximizing profits, are fairly new. This applies to countries in Western Europe rather than the US, where news production has been handled by commercial enterprises for more than a century. In Western Europe, commercial broadcasting has only been growing since the past two decades, which in turn has caused a growing scholarly interest in the commercialization of news media (McManus, 2009). Several studies show that market pressures on news organizations and its journalists to remain profitable are growing (McManus, 1995; Lewis, Williams & Franklin, 2008). The shift towards maximizing profits means that customer demand plays an increasingly
important role in the shaping of the product (McManus, 1995). McManus (1995) proposes a market-based model that shows how markets influence commercial news production. This market-model implies that “journalists should please audiences or at least those audiences that advertisers find attractive” and that “Consumer demand is the ultimate arbiter of the news product.” (Schudson, 1998: 135). For the purposes of this thesis, probably the most important relationship in the model is the one between news consumers and the news department. Put simply, in this relationship the public’s attention is exchanged for information brought by the news department (Hamilton, 2004; McManus, 1995; McManus, 2009; Strömbäck & Karlsson, 2011). The results of this relationship have a significant influence on all other relationships that are important to a news department; the size of the audience influences the fees for advertising space, it attracts sources, results in higher profits for owners/investors and therefore, a large part of the company policy is based on the expected behaviour of the consumer. The relationship between a news department and consumers proves to be of significant importance and is therefore highly interesting to research.
Scholars that study the news market argue that the behaviour of the consumer is more complex than simply investing time in news, contending that the audience often employs a fundamental cost-benefit analysis (Hamilton, 2004; McManus, 1995; Rivers et al., 1980). Before choosing a news outlet, the consumer weighs the expected rewards – which can take various forms, such as being entertained or informed - against the expected costs or efforts to gain that reward (Rivers et al., 1980). The primary cost for consumers is their time, as time spent consuming an article or news broadcast cannot be spent on other activities. This means that news outlets are not only competing against each other, but also against any alternative way for consumers to spend their time. For example, a consumer can have the option to either go for a bike ride or read the newspaper. News outlets are therefore increasingly competing in a public attention market (McManus, 1995), in which they constantly need to respond to
consumer demand. One of the ways in which news outlets attempt to gain the public’s attention and respond to consumer demand is through sensationalism. The following paragraph will work towards a definition of sensationalism.
Sensationalism
Taken from the existing literature, there does not seem to be an exhaustive definition of sensationalism. As Sparks and Tulloch (2000) point out, there are several characteristics of tabloidization and sensationalism, but not all coverage contains all characteristics. Some newspapers may have the size and shape of a tabloid, but not the content, nor the design. Furthermore, the term is often used for other media as well, such as television or radio. These media have different characteristics, functions and content mixes. Therefore it is important to form a clear overarching definition.
Grabe et al. (2001) state that since the day of the penny press, “(…) sensationalism has been defined as provoking, emotional and sensory stimulation (…)(p641)”. These characteristics can be seen as the core of sensationalism: sensationalistic news is aimed at evoking sensations in the audience. Further elaborating on how these characteristics translate to textual and visual aspects of articles, Esser (1999) states that in sensationalized texts the reader is addressed in a more jovial way, and these texts tend to contain emotional and scandalizing terms in the reporting of politics (Esser, 1999). Further exploring the latter, an example of emotionalizing or scandalizing a story is eliciting fear by stressing the threat of a situation that is reported on. In his study on newspaper reports on climate change, Höijer (2010) found that tabloids made use of terms such as “Climate Threat”, “Catastrophe”, and fear-related words such as “seriously threatened” in their reports. Furthermore, metaphors were used that elicited fear or worry. Headlines on climate change stated, “These are satellite images that reveal how sick our planet is” (Höijer, 2010), instead of a more rational headline
that would simply mention satellite images showing signs of climate change. Pictures may also elicit emotions or sensation. For example, when reporting on the Ebola outbreak in 1995, some British newspapers used images of scientists working in protective clothing, while others showed pictures of people looking fearful into the camera, or a picture of a Western nun who had died of Ebola (Joffe, 2008). These last examples also characterize the use of human-interest frames, another staple feature of sensationalistic coverage. McLachlan & Golding (2000) point out this type of frame as typical for sensationalized articles. Within the human-interest frame, journalists focus on entertainment, emotions, and especially compassion for the actors of a story. Examples of headlines of human-interest stories are “The fighting mouse who stood up to a cat” or “The horrible daffodil eaters” (McLachlan & Golding, 2000). Further adding to the textual and visual aspects of sensationalized articles, McLachlan & Golding (2000) specify greater use of easily understood illustration and simpler vocabulary, syntax and presentation to this list. Therefore, most aspects of the story are simplified, and the article stays on the surface of the information.
In conclusion, sensationalized texts are provoking, emotional and sensory stimulating. Moreover, sensationalized texts are shorter than non-sensationalized texts, there is more use of pictures, and the layout of the text includes sensory stimulation with for example flashy and entertainment-focused headlines; the vocabulary is basic and the reader is addressed in a more provocative, jovial way. Furthermore the text intends to elicit emotional reactions, and the human-interest frame is often included.
Consumer demand in political news
Now that a definition of sensationalism is formed, it is possible to examine the preference of people for this type of news, and their reasons for this preference. Evidence for some degree of a general preference for the sensationalist type of news is found in several
European countries, where news outlets admit that they are now responding to consumer demand more, and that this demand often prefers sensationalistic news. In a study by Metykova (2008) journalists from eleven European countries were asked about the effect of growing market pressures on their relationship with the audience. Journalists from all countries indicated that having a better knowledge of the audience’s preferences has become a fundamental part of this relationship. Commercial pressures, as many journalists say, drive the search for these preferences, which often turn out to be sensational (Metykova, 2008). One of the Dutch journalists summarized the situation:
“Low culture can now also be published front page once in a while. In the past you only had the acronyms of the trade unions in the headlines on the front page. We think most readers are interested in that [light news] even if they would not admit it to themselves. Research shows this. The culture of ‘we [journalists] determine what is interesting’, whole page articles about boring subjects … is gone. Today we think more about: Do people actually find it interesting? Can they understand it? Is it presented accessibly enough? The newsrooms now think about all this. This is in my view an improvement and they have to.” (Metykova, 2008: 50)
The audience’s sensational preferences have been investigated by few studies. For example, Iyengar et al. (2004) conducted a study on the audience’s preference for horse-race frames, opposed to in-depth political reporting. During elections, sensationalism often translated into more use of the horse-race frame (Cappella & Jamieson, 1996; Hamilton, 2004). Within the horse-race frame, journalists report on political elections as being a race. Articles contain short information on the latest developments in the polls and the main focus
is on which politician is winning and which one is losing at the moment (Strömback & Van Aelst, 2010; Strömback & Dimitrova, 2006; Strömback & De Vreese, 2011; De Vreese, 2005). Iyengar et al. (2004) found in their study on horse-race frames that when audiences have the choice to either read in-depth political articles or horse-race framed articles, they prefer the latter. This confirms again that responding to consumer demand often means offering sensationalist articles.
Now that the audience’s preference is established, together with the news outlets’ willingness to cater to this preference, the reasons behind this preference are sought. As mentioned before, these reasons have not yet been investigated by previous studies. Therefore, the next step is to investigate the audience and its consumer motives more closely. According to Hamilton (2004), there are three possible reasons for citizens to choose to consume political information, each of which is based on a personal demand of knowledge. The first reason is that people believe it is their duty in democracy to participate in politics. They therefore feel that they should be informed on politicians and policies, in order to make a valuable vote decision (Hamilton, 2004). The second reason is that people actively study political information because they find it implicitly interesting. They are interested in the strategies, statistics and mysterious details of politics, and like to be up-to-date on the overall political landscape (Hamilton, 2004). The third possible reason for people to value political information is their attraction to the drama that is entrenched in political races. This segment of the audience seeks entertaining stories, and is attracted to the human-interest side of the elections. If the media needs to respond to the last type of consumer demand, it leads news outlets to report in a more sensationalist style (Hamilton, 2004).
Audience motivations
In order to move on from consumer demand towards audience motivations, the Uses and Gratifications approach may provide helpful insights. When studying mass communication, the Uses and Gratifications approach (U&G) is a commonly used theory (McQuail, 1994; Ruggiero, 2000). In earlier years of communications research, scholars developed the approach to investigate the gratifications that draw and hold audiences to the types of media and content that caters to their social and psychological needs (Ruggiero, 2000). Since Hamilton (2004) already established the types of audiences, together with some of their preferences, U&G can be useful to take the next step in investigating these audience’s motivations for reading sensationalistic content. Examples of studies using the approach in these earlier years are those of Waples et al. (1940) on reading, Cantril and Allport (1935) on listening to the radio and Berelson (1949) on the functions of reading newspapers. While these traditional effect studies were mainly focused on the side of the communicator, Windahl (1981) argues that later U&G research examines mass communication from the audience’s perspective, which includes the audience’s motivations to choose a certain media outlet or type of content (Windahl, 1981). Studies on audience gratifications usually dichotomously classify media content as being essentially informational/entertaining or surveillance/entertaining. The gratifications, therefore, fall into similar categories: entertaining, surveillance and informational media uses (Katz et al., 1973). Adding to that, social utility is also an often-used category in U&G research (Ruggiero, 2000). This last gratification means that people may consume news to reinforce decisions and acquire information that can be of use in conversations (David, 2009; Ruggiero, 2000). Going back to Hamilton’s (2004) aforementioned three categories of news consumers, most of the U&G gratifications can be linked to the motivations of the three different consumers by Hamilton (2004), except the social utility gratification as this is missing from the Hamilton (2004)
typology. The U&G gratifications are summarized in table 1 below, together with the matching Hamilton (2004) motivations. By linking the gratifications to the consumer groups, the motivations of these groups to prefer sensationalistic articles are easier to investigate, since many studies on measuring these gratifications have already been done and can be used for reference.
Table 1
U&G Motivations with Matching Audience Motivations of Hamilton (2004)
U&G motivation Hamilton (2004) motivation
Informational guidance Being well-informed in order to fulfil duties as democratic citizens
Surveillance Out of interest
Entertainment Attraction to drama
Social utility -
The informational gratifications can be linked to a first group of people, who see it as their duty as democratic citizens to participate and be politically well-informed. They therefore search for information that can guide them in these democratic duties, such as factual information on parties and issues. It is therefore expected that:
H1: News consumers who have mainly informational guidance gratifications, are more likely to prefer non-sensationalized articles over sensationalized articles.
A second group of people study political news because they find it inherently interesting. This group matches the characteristics of surveillance use of media: trying to keep
up with the political landscape and satisfying their exploratory drive (Katz et al., 1973). This group is also expected to prefer in-depth political articles that do not focus on entertainment or try to elicit emotional reactions, as this is not what they are looking for:
H2: News consumers who have mainly surveillance gratifications, are more likely to prefer non-sensationalized articles over sensationalized articles.
The preferences of the third group of people match entertainment gratifications. According to Hamilton (2004), this group is attracted to drama, entertainment and human-interest stories. They would prefer entertaining articles with political scandals, flashy headlines and other sensationalistic characteristics, and therefore the following is expected:
H3: News consumers who have mainly entertainment gratifications, are more likely to prefer sensationalized articles over non-sensationalized articles.
Hamilton (2004) does not discuss the possibility of people reading political news based on social utility gratifications. Still, as mentioned before, this gratification is often used (Ruggiero, 2000), and some scholars even see it as the most relevant gratification for news use (David, 2009; Eveland et al., 2003; Rubin & Perse, 1987). Investigating the gratification of social utility could provide more insight in the audience’s motivations to prefer sensationalistic political articles, and should therefore not be overlooked in this study. Results of several studies have shown that there is a substantial relationship between anticipated social interaction and media exposure, even when the results are controlled for variables such as education level and personal interest in the studied topic (Atkin, 1972). Research on the role of the social utility gratification and media use is often focused on the frequency of media
use, rather than the type of media that is used. But as this thesis aims to discover audience motivations for a sensationalistic type of news, the link with social utility and the style of the news needs to be investigated. Pettey (1988) found that a person’s perception of his or her social environment plays a significant role when making decisions on media use. When looking at the frequency of media use, this means that the more people think conversations about political issues may be started by members of their social group, the more they will read political news articles (Pettey, 1988). When focusing on sensationalistic news in particular, this could mean that people who expect that their environment will start conversations about political drama or scandals (rather than in-depth conversations about political news) will tend to read sensationalistic news articles. The following hypothesis is formulated:
H4: People who say that they have sensation-focused conversations about politics with their environment, are more likely to prefer sensationalized articles over non-sensationalized articles.
Research Method
To test the hypotheses, a web-based survey was conducted in which respondents were asked to read one of two offered news articles on the same news issue. The respondents chose an article based on the article preview, which showed the headline, sub headline, picture (in case of the sensational article) and first paragraph (for the article previews, see appendix 1). One preview had the characteristics of a sensational news article and the other preview was more in-depth and did not have the sensational characteristics. The texts were created by the researcher, but based on actual news stories. After choosing and reading an article, respondents answered a survey questionnaire that asked them about their motivations for their
decision to choose one of the two texts. Respondents were asked to repeat the choice between a sensational and non-sensational political article three times.
Sample
The participants were Dutch adults (21 and older), as they represent 90.0 percent of the newspaper readers in The Netherlands (NOM Print Monitor, 2014). The survey was spread among 200 people through the internet, on social network sites and via email. The motivations for choosing sensationalized political news articles were measured among 85 respondents (N = 85). There were 126 recorded responses, but due to missing and wrongful answers in 41 of the responses, only 85 were useable for this study. The respondents were between 21 and 67 years old, with an average age of 32.32 years old (SD = 13.82). Of the respondents, 36.5 percent was male and 63.5 percent was female. More than half of them (54.11 percent) completed, or was still following, the highest education level in the Netherlands: a university Bachelor or Master (M = 6.47, SD = 0.62). On average, the respondents indicated that they were slightly interested in politics (M = 3.39, SD = 1.00). Of the three times that the respondents chose an article, they chose a sensationalized article once on average (M = 1.01, SD = 1.07).
As the survey required respondents to make three separate decisions between two articles, each time followed by the same survey questions on their motivations for that decision, the amount of singular tests was three times as high as the number of participants. This brings the number of singular tests to 255. This larger amount of comparable data makes it possible to conduct a more precise analysis on this group of respondents.
Operationalization and measures
The sensational texts had the aforementioned sensational characteristics, being short and simple with flashy and entertaining headlines. The reader was addressed in a jovial, provocative manner, while using emotional or scandalizing terms and metaphors. For example, one of the three the sensational texts on an Ebola outbreak read ‘ (…) the Ebola crisis has developed into a complex emergency situation that has completely shut down the struck countries’, whereas the non-sensational article read ‘(…) the Ebola crisis has developed into a complex emergency situation, with large political, social, economic and safety issues.’ Furthermore, the human-interest frame was used and pictures were included in the texts. The non-sensational text had opposite characteristics; longer, more complex words were used and the text went deeper into the political issue. The headlines were rational and sober. For example, the headline of one of the non-sensational articles was ‘United Nations start special Ebola mission Unmeer’, whereas the sensational version of the headline wrote ‘The last shred of hope: Unmeer must stop destructive Ebola virus’. Furthermore, the reader was addressed in a more distant manner than in the sensationalized text. There was no use of the human-interest frame and no pictures were included (for the articles, see appendix 2).
The motivations were measured through the survey (for the survey, see appendix 3), and were grouped into informational guidance motivations, surveillance motivations, entertainment motivations and social utility motivations using a factor analysis. These groups are based on previous similar studies on audience motivations, in which similar items were formed after factor analysis (Kaye & Johnson, 2002; Kaye & Johnson, 2004; Lin, 2006; Yang, 2004). After making their choice between the two versions and reading the articles, respondents were offered a list of statements that indicated why they might have preferred one of the two texts. The participants were then asked to indicate to what extent they agreed or disagreed with each statement, using a 5-point Likert scale, with (1) strongly disagreeing and
(5) strongly agreeing. Each statement, listed in Table 2, measures one of the four motivations. The statements are based on the same aforementioned existing survey studies on motivations for different types of media use (Kaye & Johnson, 2002; Kaye & Johnson, 2004; Lin, 2006; Yang, 2004).
Table 2
Statements measuring their matching motivation. “I chose to read this article because, based on the article preview it might...”
Informational guidance
… offer me more information on the parties involved … help me decide about important issues
… offer me unbiased viewpoints Surveillance
… help me to find out about issues that affect people like myself … help me learn what is happening in the world
… help me to keep up with the main issues of today … concern a political issue that I am interested in Entertainment
… be entertaining … offer excitement … help me relax Social utility
… give me something to talk about with others … be of use as ammunition in arguments with others … reinforce my previous thoughts on the subject
Towards the end of the survey, respondents were asked to rank the four possible motivations according to the degree of importance of each statement for their decision to choose one of the two articles. The ranking ranged from (1) most important to (4) least important. Respondents were also asked about the political conversations they have with their environment; how often they have these and if these conversations are in-depth on issues or more about political scandals.
Manipulation test
To check whether the respondents fully read and understood the articles, a manipulation test was conducted. Respondents were asked to answer eleven questions per article about the parties and persons that were mentioned and about the main focus of the article. The answers were used to compute a summated variable that measured the amount of right answers each respondent had given, with a possible minimum of zero and a possible maximum of eleven right answers. Respondents scored an average of 9.38 right answers (SD = 1.67) over all three articles, and 39.64 percent of the respondents answered all questions right. This means that over all, they fully read the articles and picked up the most important parts of the articles.
Results
Motivations
To test whether the measure taps the four types of motivations described in the theory – informational guidance, surveillance, entertainment and social utility – a principal component-analysis on the motivations reported by the respondents was conducted. The results show three components with an eigenvalue higher than 1 and the scree plot shows a
clear bend after these three components. As the previously discussed theory proposed four motivational classifications, this means that one of these classifications was not measured by the survey questions. Table 3 shows that the items that were expected to measure social utility (“give me something to talk about with others”, “be of use as ammunition in arguments with others” and “reinforce my previous thoughts on the subject”) loaded on two different factors, which means that the survey did not measure social utility as a single factor. In an attempt to repair this omission, the last social utility item (“reinforce my previous thought about the subject”), was left out. The reason for this is that the other two items are clearly social, but this item stands out, as it could also be perceived as internal. However, this attempt did not have any effect on the previous outcome. Therefore, hypothesis 4 cannot be tested. The social utility items, however, are not deleted from scales that they load on, as “give me something to talk about with others” apparently seems to measure surveillance motivations and “be of use as ammunition in arguments with others” and “reinforce my previous thoughts on the subject” seem to measure informational guidance motivations. Moreover, deleting any of these three items from the scales would result in lower scale reliabilities.
The set of items used, did measure informational guidance (eigenvalue 4.16). The variable “offer me more information on the parties involved” has the strongest association (factor loading is .81). The item “concern a political issue that I am interested in” was included in the survey to measure surveillance motivations, but appears to measure informational guidance motivations. Surveillance was also measured by the set of items (eigenvalue 2.47); here, the variable “help me learn what is happening in the world” has the strongest association (factor loading is .85). The item “offer me unbiased viewpoints” was included in the survey to measure informational guidance motivations, and the item “offer me excitement” was included to measure entertainment motivations, but these items now appear to measure surveillance motivations. Lastly, entertainment was measured (eigenvalue 1.12),
in which both variables “be entertaining” and “help me relax” associated strongly (factor loadings are .84). The three components account for 59.55 percent of the variance. All motivational statements had positive factor loadings higher than .45 on the component they scored highest on, and therefore all statements were included. Summated indexed of each factor were created by summing the individual variables, and a reliability analysis was conducted. The scales measuring surveillance (Cronbach’s alpha = .73), informational guidance (Cronbach’s alpha = .79) and entertainment (Cronbach’s alpha = .78) are all fairly reliable and their reliability could not be improved by dropping items.
Table 3
Rotated Factor Loadings of the Motivations for Choosing an Article
Factors
“I chose to read this article because, based on the article preview it might...”
1 2 3
Factor 1: Surveillance
give me something to talk about with others offer me unbiased viewpoints
help me to find out about issues that affect people like myself help me learn what is happening in the world
help me to keep up with the main issues of today
.67 .55 .53 .85 .73 .22 .05 .33 -.02 .15 .20 -.34 .16 -.16 -.33 Factor 2: Informational guidance
offer me more information on the parties involved concern a political issue that I am interested in help me decide about important issues
be of use as ammunition in arguments with others reinforce my previous thoughts on the subject offer excitement -.11 .18 .52 .40 .37 .16 .81 .74 .56 .61 .47 .57 -.10 -.03 .13 .24 .34 .46 Factor 3: Entertainment be entertaining help me relax -.02 -.19 .13 .01 .84 .84 Eigenvalue Variance explained 4.16 31.99 2.47 18.96 1.12 8.60
Figure 4 shows the respondents’ mean scores on each of the three motivation scales. On a 5-point Likerscale from 1 (strongly disagree with motivational statement) to 5 (strongly agree), respondents scored an average of 2.92 (SD = .81) on the scale measuring informational guidance. On the surveillance scale, they scored an average of 3.51 (SD = .74) and on the entertainment scale respondents scored an average of 2.01 (SD = .94). This indicates that the respondents, in general, had mostly surveillance motivations. Moreover, on average, respondents disagreed with the statements measuring entertainment motivations. This means that, overall, respondents were not driven by entertainment motivations.
Figure 4
Mean Scale Scores
Figure 5 again shows the mean scale scores, but with the choices for non-sensational articles and the choices for sensational articles measured separately. Of the 255 choices that were made, respondents preferred a sensational article 78 times and a non-sensational article
0 1 2 3 4 5
= .85) on the informational guidance motivations. Those who chose the sensational article scored an average of 3.00 (SD = .70) on the informational guidance scale. A t-test shows that this difference is not significant, t (169,26) = -1.46, p = .145, CI = [-.37, .005]. Respondents who chose the non-sensational article scored an average of 3.64 (SD = .75) on the surveillance motivations scale, those who chose the sensational article scored an average of 3.27 (SD = .65) on this scale. This difference is significant, t (221) = 3.69, p < .001, CI = [.17, .57]. Respondents who chose the non-sensational article scored an average of 1.89 (SD = .92) on the entertainment motivations, those who chose the sensational article scored an average of 2.21 (SD = .93) on this scale. This difference is significant, t (219) = -2.42, p < .05, CI = [-.58, -.05].
Figure 5
Mean Scale Scores per Article Preference
0 1 2 3 4 5
Informational guidance Surveillance Entertainment
Non-‐sensational Sensational
Article decisions and motivations
The relationships between the motivations and the preference for sensational or non-sensational news articles are the focus of this study’s research question. In the first step towards testing the hypotheses, a rank order analysis was conducted on the survey item that asked respondents to rank four motivational statements, each one representing one of the four motivation categories. Respondents ranked these statements according to their importance for their decision for one of the two articles, with value 1 being most important and value 4 being least important. The rankings were then recoded into their opposite values, making value 1 least important and value 4 most important. This would make them more intuitive to interpret, as high scorings will now mean high importance. A Spearman’s rank correlation showed no significant correlation between article preference and the entertainment or social utility motivations, but did show a significant correlation, rs = .11, p > .05, between the informational guidance motivation and article preference. The relationship is weak and positive: the more often people chose a sensational article, the more important the role of the informational guidance motivation was in that decision. This contradicts hypothesis 1.
The rank correlation also showed a significant correlation, rs = -.16, p < .05, between the scores for the surveillance motivation and article preference. The relationship is weak and negative: the more often people chose a sensational article, the less important the role of the surveillance motivation was in that decision, which is in line with hypothesis 2.
In order to further investigate whether certain motivations were more or less associated with a preference for a (non) sensational article, the motivation scales are used. A dummy variable was created that indicated whether the respondent preferred the non-sensational (value 0) or non-sensational (value 1) version of the article. A binomial logistic regression model with the preference for a sensational or non-sensational article as the dependent variable and the surveillance, informational/guidance and entertainment scale as
the independent variables, controlling for age, sex, education level and political interest, was statistically significant, x2(8) = 18.48, p > .05. The model explained 53.5 percent (Nagelkerke R2) of the variance in article preference and correctly classified 85.7 percent of cases. Table 6 shows the results of the motivation scales in the model. Significant relationships with article preference were found for informational guidance motivations and surveillance motivations, but not for entertainment motivations. This shows that, contrary to what was expected (H1), the stronger the informational guidance motivations, the higher the odds that someone chooses a sensational article. An explanation for this unexpected outcome could be found in the factor loading of the item “I chose this article because it could offer me excitement” on the informational guidance factor. This item was initially included in the survey to measure entertainment motivations, for which it was expected that people who scored high on these motivations, would prefer sensational articles (H3). The item did load positively on the entertainment factor, but lower than on the informational guidance factor. In an attempt to explain the outcome of hypothesis 1, this item was deleted from the scale, which left a fairly reliable scale with a Cronbach’s alpha of .77. This is just slightly lower than the reliability of the original informational guidance scale (Cronbach’s alpha = .79). Running the logistic regression with this new informational guidance scale did not, however, change the outcome of hypothesis 1. Hypothesis 1 is therefore rejected. In an attempt to further explore this unexpected outcome, the open answers respondents gave on the question why they chose one of the two news articles are investigated. Many answers were in line with the expectations: people with high scores on the informational guidance scale who preferred the non-sensational articles, often referred to their preference for an objective and factual writing style in their answers. Others answered the question by pointing out what they did not like about the sensational article; the title was too sensational or emotional for their likings. However, other answers contradicted the expectations. Respondents who scored high on the
informational guidance scale and preferred the sensationalized article, sometimes gave interesting explanations for their preference; they viewed the sensational characteristics as positive cues that indicated more information, easier to understand articles and that issues would be taken more seriously:
“Because there were more details in the text, I could understand it a lot quicker.”
“The picture gave me more clarity on the subject.”
“The headline is more catchy, which makes me believe that the issue will be taken more seriously.”
“The preview looked like it would give me more information about ebola that I don’t know yet.”
“It offered me information on recent developments in The Netherlands that seemed interesting to me.”
The answers above show that these respondents were clearly driven by informational guidance motivations, searching for serious, in-depth information, but they looked for the sensational cues in the articles to provide this type of information, and therefore preferred the sensational news article.
The regression model does confirm hypothesis 2 (H2), showing that the stronger the surveillance motivations, the lower the odds that someone chooses a sensational article.
Table 6
Results of Logistic Regression Predicting Article Preference
B SE B eB
Informational guidance 1.99** .87 7.34
Surveillance -3.43*** .93 .03
Entertainment - .34 .47 .71
Note: Controls are age, sex, education level and political interest (omitted from the table). eB = exponentiated B. **p < .05. ***p < .001.
In order to attain more insight into the size of the significant relationships, the logistic regression model is further analyzed, by calculating the probability margins of the two relationships. The outcomes indicated that for respondents who had low scores on the informational guidance motivations, the odds of choosing a sensational article over a non-sensational article were 5.5 percent. For respondents with high scores on the informational guidance motivations, these odds were 77.0 percent. For respondents with low scores on the surveillance motivations, the probability of choosing a sensational article over a non-sensational article was 92.5 percent. For respondents with high scores on these motivations, the probability was only 5.0 percent.
Conclusion and discussion
The aim of this study is to expound on our understanding of people’s motivations for preferring sensational political news articles over non-sensational political news articles. In order to achieve this, four types of motivations – informational guidance, surveillance, entertainment and social utility – are investigated. The results reject one of four hypotheses posed; people who are led by informational guidance motivations prefer sensational news
articles over non-sensational news articles. Hypothesis 2 is accepted, as the results show that people who are led by surveillance motivations prefer non-sensational political news articles. There were no significant results found for the relationship between article preference and the entertainment motivations. Furthermore, it was not possible to test the social utility hypothesis.
Evidence presented here suggests that, contrary to the expectations, individuals who read a political news article to find information that can help them define their position on certain political issues and parties (informational guidance motivations), are more likely to prefer sensationalized news articles. This rejects our expectations posed by hypothesis 1 (H1). The answers to the open ended question that asked the respondents why they chose on of the two news articles, indicated a conscious preference for the sensational article with people who are driven by informational guidance motivations, mainly because they perceived the sensationalized texts as easier to understand and more accessible. Looking back at Hamilton’s (2004) description of the informational guidance audience, it now makes sense that they might want to acquire objective information in an easy way, as they are just looking for information to fulfil their democratic duties. This suggests that a sensationalized writing style could actually cater to a segment of the public that is interested in politics, but prefers a more accessible writing style, which supports the idea of an alternative public sphere that engages different audiences in politics (Örnebring, 2007; Örnebring & Jönsson, 2007).
Our second hypothesis (H2) is confirmed in twofold by the results presented. The group of people Hamilton (2004) describes as being motivated by interest, wanting to keep up with developments in the political landscape and the main political issues that may affect them as a person, are more likely to prefer a non-sensational news article. Zooming in on one of the core surveillance motivations – keeping up the main issues of the day – the evidence
shows that people with a higher preference for sensational articles are less motivated by this motivation.
There was no significant relationship found between the entertainment motivations and article preference (H3), which is surprising, considering that entertainment is a core aspect of sensationalism. The results show that, overall, respondents were not driven by entertainment motivations in their article choices. This explains why there was no significant relationship found. However, more important is the question why the respondents were not driven by entertainment motivations. An answer could be found in the composition of the sample, as the sample demographics showed that the respondents were mostly high-educated people. Previous studies have shown a negative relationship between education level and the preference for political news articles with entertaining characteristics (Prior, 2003; 2005). Therefore, the composition of the sample possibly played a large role in the rejection of hypothesis 3.
Furthermore, the attempt to create a scale that measures social utility failed. Looking back at the theory, this is in line with Hamilton’s (2004) theory that distinguishes three types of audiences, from which not one is driven by social utility motivations. It also demonstrates the earlier notion that though this type of motivations is often used, and by some deemed highly important in audience motivation studies, it has not been recognised as such by all scholars. The uncertainty around the social utility motivations is again illustrated by the way Kaye and Johnson (2002; 2004) measure these. The items that were used to measure social utility in the present study were taken from Kaye and Johnson’s (2002; 2004) previous studies on audience motivations for using the internet as a source of political news. In their 2002 study, Kaye and Johnson measure social utility with the same items that were used here. These items are focused on finding information that can be used in conversations or discussions with others. But in their 2004 study, Kaye and Johnson mixed the social utility
items with entertainments items, and even added a surveillance item to the scale. The lack of clarity around social utility motivations requests further research in this field of study.
There are several limitations to this study, which mainly relate to the sample. Because of limitations in time and means, the amount of respondents that were reached was quite small, which makes it more difficult to find significant relationships, and to project these relationships onto larger groups of people. The respondents’ demographics add to this rather low external validity; as most respondents were young, high-educated women, they were not representative of the entire scope of Dutch citizens. Furthermore, the large amount of survey data that was not useable, or just partly useable, because large numbers of questions were left unanswered, could have been reduced by making it impossible for respondents to skip to the next page in the survey without answering all questions.
In conclusion, this study has provided insight into the motivations behind sensationalistic audience demand. Though this subject requires further investigations, the results show possibilities for sensationalized political news in democracy, reaching a segment of the audience that is interested in more easily accessible political information.
Appendix 1 Article Previews
Article 1a (sensational) Article 1b (non-sensational)
Appendix 2 Articles
Article 1a (sensational)
The last shred of hope: Unmeer must stop destructive Ebola virus
The United Nations has taken on a special mission to save those countries struck most by the wildly spreading virus that does not only leave people dead, but also in fear.
The UN wants to start the mission, called Unmeer, this month. The main goals are ending the outbreak, treating infected people and bringing stability in Guinea, Sierra Leone and Liberia.
UN’s Secretary–General Ban Ki-moon stated that the Ebola crisis has developed into a complex emergency situation that completely shut down the struck countries.
The healthcare system in Liberia, for instance, is completely shut down. According to Ban, this has lead to more people dying from ‘normal’ diseases than from the incurable Ebola. The food prices are rising (for the often already poor citizens) and public services are disrupted.
Boarders
This Thursday, the fifteen countries in the council took on the mission unanimously, agreeing to provide aid and means in the battle against the wildly spreading virus.
The Security Council calls on countries to no longer close their boarders for those countries that have been struck by the virus. This would not help the fight.
The mission is meant to enable a “fast mobilisation” of people, materials and financial means.
Life standing still
Life in the West African country Sierra Leone will be standing still for three days straight, from midnight on. The six million inhabitants will have to stay inside their homes while volunteers go from door to door to track infected people who are hesitant or refuse to get treated. The volunteers will also hand out soap and information on the disease.
Many people who are possibly infected are scared. They do not want to visit a hospital because they believe that people just die there. There have also been cases where dying patients were simply refused when clinics were full.
Next to that, the fear and confusion about the terrifying disease have caused a lot of unrest. In Guinea a completely innocent disinfection team has gone missing after they were attacked by a group of youngsters.
There is no cure for Ebola. The disease can only be controlled by isolating patients. The virus is transferred through body fluids and especially aid workers risk being infected.
Article 1b (non-sensational)
United Nations start special Ebola mission Unmeer
The United Nations has taken on a special mission to fight Ebola in the most heavily struck countries Liberia, Guinea and Sierra Leone.
The UN wants to start the mission, called Unmeer, this month. Different tasks will be carried out, such as ending the outbreak, treating the infected and bringing stability. Adding to that, ‘essential services’ need to be assured and further outbreaks need to be prevented.
UN’s Secretary-General Ban Ki-moon stated that the Ebola crisis has developed into a complex emergency situation, with large political, social, economic and safety issues.
Boarders
The UN Security Council labels the outbreak as a “threat to the international peace and safety”. This Thursday, the fifteen countries in the council took on the mission unanimously. The statement calls on all countries to provide needs and assistance in the battle against the virus.
The Security Council also calls on countries to no longer close their boarders for the countries that have been struck by the virus. This contributes to the isolation of the struck countries and could possibly hinder the control of the virus, according to the statement.
Secretary-General Ban Ki-moon announces he will appoint a special ambassador for the UN-mission Unmeer.
Lock-down
The six million inhabitants of the West-African country Sierra Leone will have to stay in their homes for three consecutive day, while volunteers will go from door to door to find infected people who are hesitant or refuse to get treatment. They will also hand out soap and
information on the disease.
This way, authorities expect to uncover hundreds of new cases of Ebola.
The Ebola outbreak in West Africa has taken over 2.600 lives. The outbreak started in Guinea, but most casualties were in Liberia. People have also died in Sierra Leone, Nigeria and Senegal.
There is no cure for Ebola. The disease can only be controlled by isolating patients. The virus is transferred through body fluids and especially aid workers risk being infected.
Article 2a (sensational)
A furious Asscher (PvdA) SLAMS the PVV for reprehensible
remarks about the Islam.
According to minister Asscher (Social Affairs),
the PVV has crossed the line by suggesting that “Muslims do not have the same rights as
other Dutch citizens”.
Minister Asscher is furious about the comments made by PVV member Machiel de Graaf on Dutch Islamic citizens. De Graaf said, among other things, that all mosques should be closed.
During the budget meeting that took place earlier this week, De Graaf called the Islam “a state-undermining totalitarian ideology that has declared war on the free West.”
Party leader Geert Wilders and De Graaf add to that, that no form of Islamic representation has the right to enter the Ridderzaal.
Asscher responded to these remarks in disgust and called them “reprehensible”. He said that the comments go against the Dutch constitution.
“Through the uterus”
During last Wednesday’s budget meeting, De Graaf bashed the Islam in The Netherlands. He said that according to some Islamic leaders, The Netherlands is being conquered “through the uterus”. “The Dutch individuality, ideology and culture are being washed out via immigration and the uterus”, he said.
Several shocked Members of Parliament directly accused him of also painting young, innocent children as the problem and as guilty. These children are Islamic, but can they be held accountable for these nasty accusations? The Graaf, however, says that the anger should be pointed at the Arabic leaders that make such remarks, such as the former Algerian
president Houari Boumédienne.
Article 2b (non-sensational)
Asscher (PvdA) responds to the PVV’s remarks about the Islam.
According to minister Asscher (Social Affairs), the PVV has crossed the line by saying that “Muslims do not have the same rights as other Dutch citizens”.
Minister Lodewijk Asscher said that last Thursday during the budget meeting. He calls the comments, which were made by PVV member Machiel de Graaf, “reprehensible”.
Asscher articulated his aversion towards the statement that all mosques should be closed. “It is not possible that a country with this constitution, is a country where all mosques are being closed and the boarders are closing for people from Islamic countries”, said the minister.
He also criticized earlier comments made by the PVV that stated that there is no place for the Islam in the Ridderzaal.
“Mister De Graaf pretends to know all kinds of facts, but has no clue of what should be most important in this house: our constitution”, Asscher said.
Ridderzaal
Last week Asscher met with several different religious organizations in the Ridderzaal. He did this because of the growing tensions between different groups in society.
The PVV responded to this event by asking to no longer allow the Islam, which they call “a state-undermining totalitarian ideology that has declared war on the free West”, into the Ridderzaal.
Party leader Geert Wilders and De Graaf state that no form of Islamic representation has the right to enter the Ridderzaal.
The PVV members stress that this is important because the Declaration of Independence of The Netherlands is placed in the Ridderzaal. This represents the “strive for freedom”, says the PVV.
“With this, the PVV goes against the essence of our democratic constitution”, Asscher says.
“Discrimination against religion, political beliefs, race, gender or on what ground whatsoever, is prohibited. That is stated in the constitution that everybody in this country has to abide to.”
De Graaf responded that Asscher should lose his “naivety”.
“Through the uterus”
During last Wednesday’s budget meeting, De Graaf attacked the Islam in The Netherlands. He said that according to some Islamic leaders, The Netherlands is being conquered “through
the uterus”. “The Dutch individuality, ideology and culture are being washed out via immigration and the uterus”, he said.
Several shocked Members of Parliament directly accused him of implying that young children are also “guilty”. The Graaf, however, says that the anger should be pointed at the Arabic leaders that make such remarks, such as the former Algerian president Houari Boumédienne.
Article 3a (sensational)
“Mommy behind bars: who will take care of me now?”
The PvdA again raises the alarm for children that are forced to live a lonely and neglected life, because their mother is in jail.
Two years ago, the party’s proposition to appoint a social worker that organizes the care of motherless children in every municipality, found support from all other parties in Den Haag. But now, sadly, these kinds of social workers are nowhere to be found. The PvdA calls it a disgrace.
Every year, nearly 3000 women are sent to jail. 70 percent of these women is a mother of at least two children, that are being left behind without their mom.
“Outright shocking”, says PvdA member Marith Rebel-Volp about the situation. She calls for a debate.
"When your mother goes behind bars, that is a very drastic event for a child. The government has the task to take care of these children, make sure that they have a roof over their heads and help with the difficult questions these children struggle with, even after their mother has returned home”, says Rebel-Volp.
Research
The PvdA carried out a study, from which the results show that children of locked up mothers were left without help by most municipals. The largest part of the municipals did not have any clue how to handle the critical situation and in many cases people were referred to the wrong organizations
Article 3b (non-sensational)
“Child of detained mother needs help”
The PvdA again raises the alarm for children who are being taken care of badly, because their mother is in prison.
Two years ago, the party’s proposition to appoint a social worker that organizes the care of motherless children in every municipality, found support from all other parties in Den Haag. But according to the PvdA, research in 55 municipals shows that only two cities in the Netherlands offer such a social worker.
Party member Marith Rebel-Volp wants a debate: "When your mother goes to prison, that is a very drastic event for a child. The government has the task to take care of these children,
make sure that they have a roof over their heads and help with the difficult questions these children struggle with, even after their mother has returned home”, says Rebel-Volp. Research
De PvdA carried out its own independent study. "It appears that, after two years, only two municipals referred to the coordinator aftercare, the appointed person that helps the children that are having issues with their mother going to prison.
"Eleven municipals did not have a clue how to handle such a situation, thirteen referred to social services, six to the centre of Youth and Family, ten to childcare and thirteen municipals referred to another organization. One time, people were referred to Service Work and Income. That is just not ok.”
Appendix 3
Survey Questionnaire
Introduction
You are invited to participate in this study on the media use of Dutch citizens, which is being carried out under the auspices of the Graduate School of Communication, a part of the University of Amsterdam.
During this survey you will be asked to read three news articles and answer several questions about these articles. The survey will take about 15 minutes of your time.
As this study is being carried out under responsibility of the ASCoR, University of Amsterdam, we can guarantee that:
1) Your anonymity will be safeguarded, and that your personal information will not be passed on to third parties under any conditions, unless you first give your express permission for this. 2) You can refuse to participate in the research or cut short your participation without having to
give a reason for doing so. You also have up to 24 ours after participating to withdraw your permission to allow your answers or data to be used in the research.
3) Participating in the research will not entail your being subjected to any appreciable risk or discomfort, the researchers will not deliberately mislead you, and you will not be exposed to any explicitly offensive material.
4) No later than five months after the conclusion of the research, we will be able to provide you with a research report that explains the general results of the research.
For more information about the research and the invitation to participate, you are welcome to contact the project leader Lola Lindhout (lola_lindhout@hotmail.com) at any time.
Should you have any complaints or comments about the course of the research and the procedures it involves as a consequence of your participation in this research, you can contact the designated member of the Ethics Committee representing ASCoR, at the following address: ASCoR Secretariat, Ethics Committee, University of Amsterdam, Postbus 15793, 1001 NG Amsterdam; 020-525 3680; ascor-secr-fmg@uva.nl.
Any complaints or comments will be treated in the strictest confidence.
We hope we have provided you with sufficient information. We would like to take this opportunity to thank you in advance for your assistance with this research, which we greatly appreciate.
Kind regards, Lola Lindhout
… I understand the text above and agree with participating in this study.
On the following page, you will be presented with a choice between two political articles on the same topic. Please click on the article that you prefer to read, and then read the article in its entirety. When you are finished, you can click “next” to continue.
Part 1
You have just read a short article on (subject). Please enter below why you chose to read this article. “I chose to read this article because…”
Q1 ………..
Please indicate to what extent you agree with the following statements: “I chose to read this article because, based on the article preview it might...”
S1. … offer me more information on the parties involved
(very unimportant) 1 2 3 4 5 (very important)
S2. … help me decide about important issues
1 2 3 4 5
S3. … offer me unbiased viewpoints
1 2 3 4 5
S4. … help me to find out about issues that affect people like myself
1 2 3 4 5
S5. … help me learn what is happening in the world
S6. … be entertaining
1 2 3 4 5
S7. … offer excitement
1 2 3 4 5
S8. … give me something to talk about with others
1 2 3 4 5
S9. … be of use as ammunition in arguments with others
1 2 3 4 5
S10. … help me relax
1 2 3 4 5
S11. … remind me of my previous thoughts on the subject
1 2 3 4 5
S12. … concern a political issue that I am interested in
1 2 3 4 5
S13. … help me to keep up with the main issues of today
Part 2
Please rank the following motivations in order of how big their role was in deciding to choose the article you have just read (1 being the largest, 4 playing the smallest role in your decision):
… This article could provide me with helpful information that I could use to form an opinion on policy issues or vote decisions.
… This article could provide me with some entertainment
… This article could provide me with information I find interesting, and to keep up with what is happening in the world.
... This article could provide me with information I could later use in conversations
Part 3
Please indicate to what extent you agree with the following statements:
S14. “I talk about politics with my friends, family or colleagues”
(never) 1 2 3 4 5 (always)
! Unless participants answer (1)- never, they will be directed to S15 (if they answer ‘never’, they will be directed to S20)
“When I talk about politics with my friends, family or colleagues…”
S15. … we talk about complex political issues (such as healthcare, international conflict, economic reforms)
S16. … we like to “keep it light” (by talking about small, or local issues, and keeping it on the surface of information)
1 2 3 4 5
S17. … we talk about political scandals
1 2 3 4 5
S18. … we talk about how we feel about certain politicians
1 2 3 4 5
Part 4
Please indicate to what degree you agree with the following statements about the texts you have just read:
“This article was…”
S19 … in-depth
(strongly agree) 1 2 3 4 5 (strongly disagree)
S20 … superficial
1 2 3 4 5
S21. … entertaining
S22. … serious
1 2 3 4 5
S23. … difficult to understand
1 2 3 4 5
S24. “The headline of this article was flashy”
1 2 3 4 5
