The Influence of Sensationalist Features in Television News Stories

Studies in Communication Sciences 17.2 (2017), pp. 183–194

The influence of sensationalist features in television news stories on perceived news quality and perceived sensationalism of viewers in different age groups Mariska Kleemans, Radboud University, Behavioural Science Institute* Paul G. J. Hendriks Vettehen, Radboud University, Behavioural Science Institute Johannes W. J. Beentjes, University of Amsterdam, Amsterdam School of Communication Research Rob Eisinga, Radboud University, RadboudSCR *Corresponding author: [email protected]

Abstract This study investigates whether the decreased trust in news media can be explained by the increase in sensationalism in news. To this end, an experiment was conducted in which viewers (N = 288) evaluated sensationalist versus non-sensationalist television news in terms of perceived news quality and perceived sensationalism. Each participant watched four television news stories that varied in the presence of two categories of sensationalism: arousing content and tabloid packaging. Findings showed that sensationalist television news tends to be more negatively evaluated than non-sensationalist news. In addition, critical views on arousing content appeared to be particularly visible among young and middle-aged adults. These findings suggest that the rise of sensationalist news could be an explanation of the declining trust in news media that is witnessed in a number of countries.

Keywords sensationalism, arousing content, tabloid packaging, news performance

1 Introduction (e.g., Bennett, Rhine, Flickinger, & Bennett, 1999; Jones, 2004, 2005). Second, the de- Public evaluations of news performance cline in evaluations of news performance have declined steadily over the past de- has been explained by the rise of Internet cades. In the United States, ratings of the and non-mainstream news outlets that press have become dramatically more offer political views that differ from those negative. For instance, people offer more offered by traditional news media. Con- critical evaluations of the news in terms of suming online information that argues its accuracy, fairness, and independence with and is critical of traditional news me- (The Pew Research Center, 2013). Com- dia might lead to declining perceptions of parable trends were found in European news performance for these media (e.g., countries. As a result, in 2016 only half of Jones, 2005; Tsfati & Ariely, 2013). A third the Europeans had trust in information explanation focuses on the role of tradi- provided by television, whereas trust in tional media, particularly television jour- the written press (46%) and Internet (36%) nalism. According to this explanation, was even lower (Eurobarometer, 2016). the profit-driven tendency in television In the literature, explanations for the journalism to paint a cynical picture of decline in the public’s evaluation of news politics by focusing on political scandals, performance have primarily been cast in conflicts, incivility in debates, and horse- the context of political communication. race news may not only have affected trust First, the decline has been conceptualized in politics (the much debated “videomal- as being part of a broader tendency to dis- aise” hypothesis, e.g., Bennett et al., 1999; trust institutions, experts, and politicians Habermas, 2006; Mutz & Reeves, 2005;

https://doi.org/10.24434/j.scoms.2017.02.004 © 2018, the authors. This work is licensed under the “Creative Commons Attribution – NonCommercial – NoDerivatives 4.0 International” license (CC BY-NC-ND 4.0). 184 Kleemans et al. / Studies in Communication Sciences 17.2 (2017), pp. 183–194

Robinson, 1976) but also trust in the news ical communications literature described media themselves (cf. Cho, Shah, Nah, & above (the label “videomalaise,” e.g., Cho Brossard, 2009). et al., 2009; Mutz & Reeves 2005). The present study is particularly re- In all, the current study aims to inves- lated to this latter explanation in that it tigate whether the decreased trust in news also investigates the impact of profit-based media can be explained by the increase news coverage on news performance. in sensationalism in television news. An However, it goes beyond profit-driven cov- experiment will be conducted in which erage of “political” news in the sense of participants of varying ages evaluate sen- news about politics or politicians. Instead, sationalist versus non-sensationalist tele- it examines a broader range of techniques vision news. We take the moderating role that can be used to attract the viewer’s of age into account, because studies have attention to a wide range of news topics, already revealed differences between older that is, the use of sensationalism. Today’s and younger viewers in their preference news stories unmistakably display an in- for sensationalist versus non-sensation- creasing presence of sensationalist char- alist news (Kleemans, Hendriks Vettehen, acteristics (cf. Hendriks Vettehen, Beent- Beentjes, & Eisinga, 2012; Lang et al., jes, Nuijten, & Peeters, 2011; Schaefer & 2005). Therefore, this study investigates Martinez III, 2009; Slattery, Doremus, & whether they also have different percep- Marcus, 2001). The question is whether tions of such news. there could be a connection between the rise of sensationalist news and the general decline in the news consumer’s evaluation 2 Sensationalism in news: arousing of news performance. News critics often content and tabloid packaging blame today’s news for being sensational, exaggerated, inaccurate, unbalanced, and The presence of sensationalist features in just looking for anything that thrills (e.g., news has already been traced back to the Jones, 2009). However, it is largely un- 1500s (Stephens, in Grabe Zhou, & Bar- known whether this opinion is shared by nett, 2001). It has been fiercely criticized viewers, and consequently, whether the with regard to, for instance, the Penny increase in sensationalist news might be Press (cf. Emery, 1962) and Yellow Jour- an additional explanation of the waning nalism (cf. Baran & Davis, 2012). Never- confidence in the news media. For this theless, sensationalist features never dis- reason, the current study investigates appeared completely from news stories, how viewers perceive the performance of and trend studies on television news even the two categories of sensationalist news show that sensationalism increased over characteristics that are commonly distin- the past decades (e.g., Hendriks Vettehen guished in the literature: arousing content et al., 2011; Hvitfelt, 1994; Slattery et al., and tabloid packaging. 2001; Slattery & Hakanen, 1994; Schaefer & We will limit our study to television Martinez III, 2009). Consequently, schol- news, because television news still is a ma- ars pay increasing attention to the causes jor source of news for many citizens in, for and consequences of sensationalist news. instance, the US (cf. The Pew Research Cen- In the literature, a number of related ter, 2013) and in Europe (Eurobarometer, concepts have been employed to study the 2016) – including the Netherlands (Sociaal general phenomenon we refer to as “sen- Cultureel Planbureau, 2015), where the sationalist news,” including infotainment current study is conducted. Moreover, (Graber, 1994), bad news (Newhagen & television news is relevant to study here, Reeves, 1992), soft news (Baum, 2002), tab- because particularly the relatively wide loidization (Bek, 2004), and arousing news opportunities provided by television’s (Kleemans et al., 2012). A number of these audiovisual character to provide a sensa- studies have defined sensationalist news tionalist coverage of events have been re- as news including content and packaging lated to audience evaluations in the polit- characteristics that have the capability to Kleemans et al. / Studies in Communication Sciences 17.2 (2017), pp. 183–194 185 elicit arousal and attention responses in tionalism (Hollifield, 2006; McQuail, 1992; viewers (Grabe, Lang, & Zhao, 2003; Grabe Schatz & Schulz, 1992). et al., 2001; Hendriks Vettehen, Nuijten, Differences in grouping of criteria into & Peters, 2008; Lang, Bolls, Potter, & Ka- dimensions have been proposed in past wahara, 1999). Regarding content, espe- research. For instance, based on core val- cially survival-relevant stimuli have been ues of the western civilization, McQuail regarded as prominent examples. Humans (1992) discerned the performance dimen- are evolutionary predisposed to survey sions of freedom, equality, and cohesion. their environment for potential threats (cf. By contrast, Schatz and Schulz (1992) Davis & McLeod, 2003; Shoemaker, 1996). discerned five dimensions: diversity, rel- Stimuli in news that refer to such threats evance, professionalism, legitimacy, and therefore attract audiences and elicit public acceptance. The present study does arousal. Examples are stories and pictures not seek to take position in questions con- referring to crimes, violence, riots, fires, cerning the precise conceptualization of and disasters (Adams, 1978; Grabe et al., the concept of media performance. In- 2001; Newhagen & Reeves, 1992; Slattery & stead, a number of indicators of media Hakanen, 1994). Following Grabe and col- performance have been included in the leagues (Grabe, Zhou, Lang, & Bolls, 2000, current study that (a) particularly relate to Grabe et al., 2003) we label this category of the performance of television news, and sensationalism arousing content. (b) have been used in earlier studies on the Packaging characteristics of news sto- performance of sensationalist news. ries that may be considered sensational As far as known, only a few studies are features that represent change or un- investigated whether sensationalist con- expected occurrences, because they also tent and packaging of news affected per- elicit attention and arousal responses in ceptions of TV news performance. In one viewers. For instance, fast pacing, sound study (Grabe et al., 2000), viewers watched effects, music onsets, and decorative ed- a number of news stories in either a stan- iting techniques such as wipes, fades, and dard or a tabloid version and subsequently dissolves are discerned in this regard (cf. answered a number of evaluative ques- Grabe et al., 2000; Grabe et al., 2003; Hen- tions about these stories. Results showed driks Vettehen, Nuijten, & Beentjes, 2005; that participants perceived the standard Lang, 2000; Lang et al., 1999). Again fol- packaged stories as more informative and lowing Grabe and colleagues, we label this believable. The other study (Grabe et al., category of sensationalism tabloid pack- 2003) had a similar design, but this time aging (cf. Grabe et al., 2000, 2003). both the content of news stories (arousing versus calm) and packaging of the stories 2.1 Assessing the performance of (standard versus tabloid) varied. Results sensationalist news showed that participants perceived the The concept of media performance refers tabloid stories as less believable and less to the assessment of mass media using objective than the standard stories, but a range of evaluative criteria (McQuail, not as less informative. Arousing content 1992). However, media performance is a stories were perceived as less informative somewhat difficult concept, which has compared to calm content stories, but not been linked to varying criteria and dimen- as less objective or less believable. In sum, sions. Objectivity is perhaps the most fre- the two studies provide a preliminary basis quently mentioned criterion in assessing to expect that viewers are somewhat skep- news media performance. Other criteria tical towards the journalistic performance that have often been applied to news are of sensationalist news. accuracy and completeness, reliability, The findings from a correlational study credibility, factualness, fairness, balance, conducted by Wang and Cohen (2009) pro- believability, informativeness, compre- vide further support for this expectation. hensive coverage, presentation of multiple This study showed that respondents who points of view, and avoidance of sensa- estimated the numbers of audio, visual, 186 Kleemans et al. / Studies in Communication Sciences 17.2 (2017), pp. 183–194 and editing techniques included in tele more at ease with digital technologies vision news as being higher, also perceived than older people (Sundar, 2008). When the news as being more sensationalist. they acknowledge technologies as being Perceived sensationalism was measured “hip” or when they are impressed by the in their study by asking respondents to in- novelty of the technology, they may eval- dicate whether eight concepts (accuracy, uate the message as more credible. As responsibility, importance, credibility, Sundar (2008) states: “If it’s cool, it’s cred- professionalism, arousing viewer interest, ible” (p. 82). Moreover, young media us- invading privacy, and gossiping) were in- ers may be enamored by the novelty and herent in the notion of sensationalism. therefore perceive such content as being The latter three concepts were reversed in more credible. In addition, the experience the analysis, so that a higher score on the that young people have with complex me- constructed scale referred to a higher per- dia interfaces may imply that they have a ception of sensationalism. Because items higher threshold for cognitive overload or such as the accuracy, credibility, and im- overstimulation than older people. There- portance of a story were included in the fore, young people are expected to have scale, their results imply that more tabloid more positive evaluations of new digital packaging features lead to more negative formats compared to older people. perceptions of journalistic performance. A similar argument can be built with The current study extends past re- regard to arousing news content and tab- search by measuring a greater variety of loid packaging. Younger viewers have indicators of news media performance. grown up in a highly arousing media en- We will include perceptions of reliability, vironment (e.g., video games, fast-paced perceptions of news performance as used music clips, and movies), while older by Grabe and colleagues, and several in- viewers will have experienced an increase dicators of perceived sensationalism, the in arousing characteristics in media mes- central concept in the Wang and Cohen sages over the years (cf. Tapscott, 2009). study. Following the general findings in Due to this generational difference, young previous research, we pose the following news viewers are more at ease with the hypotheses: increase in arousing content and tabloid produced news. We can thus expect that H1a: Viewers perceive the performance younger viewers are more positive towards of arousing content news stories as the performance of arousing content and more negative than non-arousing con- tabloid packaged stories than older view- tent news stories ers. We therefore expect that:

H1b: Viewers perceive the performance H2a: Older viewers have more negative of tabloid produced news stories as perceptions regarding the media per- more negative than standard produced formance of arousing content (as op- news stories posed to non-arousing content) compared to younger viewers 2.2 Age differences in perceptions of sensationalist news H2b: Older viewers have more negative No previous research has investigated the perceptions regarding the media perfor- moderating role of age, although there are mance of tabloid produced news stories indications that sensationalist news is par- (as opposed to standard produced news ticularly salient to young people and that stories) compared to younger viewers young people will therefore have more positive evaluations compared to older people. With regard to digital technologies, Sundar (2008) argues in his MAIN Model that the current generation is more eager to try new formats and is, therefore, Kleemans et al. / Studies in Communication Sciences 17.2 (2017), pp. 183–194 187

3 Material and methods The dataset of the content analysis study, containing statistics regarding the pres- 3.1 Stimulus materials ence of sensationalist content and packag- An experiment was conducted in The ing features in a large number of television Netherlands as part of a larger project on news stories, was used to select stories that the effects of sensationalist news (cf. Kl- were potentially useful for the experiment. eemans et al., 2012, Kleemans, Hendriks The rationale behind the selection process Vettehen, Eisinga, & Beentjes, 2014) using was that we aimed to maximize the differ- 16 Dutch public service news stories as ence between the two types of content and stimulus materials. All stories had been the two types of packaging in the stimulus broadcast in the main television news materials. Therefore, we first selected only program in The Netherlands, the NOS stories that included either an arousing Journaal. The stories were selected out of story topic combined with arousing pic- a pool of news stories that had been con- tures or a non-arousing story topic with- tent analyzed in a previous study on the out the presence of any arousing pictures. presence of arousing content and tabloid Second, from this sub-selection we select- packaging features in news stories (cf. ed the most suitable stories based on the Hendriks Vettehen et al., 2012). As shown presence or absence of tabloid packaging in Table 1, the intercoder reliabilities of this features: stories including either a very content analysis were generally acceptable small (i.e., a low number of camera chang- (i.e., above .67) or good (i.e., above .80; es and no sound effects, music onsets or Krippendorff, 2004). Exceptions were the decorative editing techniques) or a very reliabilities for the “wipes” and “dissolves,” large number of tabloid features were se- which where slightly lower, but these can lected. be explained by the relatively small sample As a final step in the selection process, size and the skewed distribution of these we selected eight stories out of the pool of variables (Zhao, Liu, & Deng, 2012). remaining stories that included all arousing content indicators and eight stories that included none of them. A number of Table 1: Coding overview of arousing additional selection criteria were applied characteristics in news to make the final choices, because we wanted to minimize the potential influ- Measures Description (Krippendorff’s alpha for ence of factors other than differences in intercoder reliabilities) content and packaging. In particular, we Arousing content only selected stories that presented lo- Arousing topic The presence of at least one of the catego- cal news, had a comparable story length ries violence, criminality, accident/disaster, (M = 95.31; SD = 10.78, measured in sec- extremism, unnatural death, sex, drugs, celebrities, and scandal (α = .78) onds), and did not include production er- Arousing pictures The number of camera shots in which rors or a slip of the tongue by the reporter. pictures of least one of the categories In addition, topics that had received a lot violence, accident/disaster, people with of media attention or controversial top- physical injuries, death per minute were ics were not incorporated in the stimulus present (α = .91) material. We also included only stories Tabloid packaging that were not particularly relevant for the Camera shots The number of camera shots per minute in participants. For instance, we selected sto- the news story (α = .97) ries that took place in other parts of the Music onsets The number of music onsets in the news story (α = .82) country than where the participants lived. Wipes The number of wipes in the news story Finally, only stories that could be matched (α = .65) based on story topic remained for the ex- Fades The number of fades in the news story periment. As described in Table 2, there (α = 1.00) thus was a standard produced and a tab- Dissolves The number of dissolves in the news story loid produced story for each topic. (α = .60) 188 Kleemans et al. / Studies in Communication Sciences 17.2 (2017), pp. 183–194

Table 2: Overview of the 16 news stories

Arousing Content Non-arousing Content Tabloid Packaging Standard Packaging Tabloid Packaging Standard Packaging Topic: Fire Topic: Fire Topic: Economy Topic: Economy Story about a fire in a museum Story about a fire in a house Story about the economic Story about increase of goods situation in The Netherlands traffic Port of Rotterdam Topic: Riots Topic: Riots Topic: Nature Topic: Nature Story about riots at the turn of Story about riots during high Story about mosquitos on Dutch Story about buffaloes in Dutch the year school demonstration Island national parks Topic: Accident Topic: Accident Topic: Culture Topic: Culture Story about an accident during a Story about an explosion in Story about 40 years public Story about Dutch buildings reg- demonstration a cafe transport (metro) istered as national monuments Topic: Crime Topic: Crime Topic: Environment Topic: Environment Story about a criminal accused Story about a pupil who was Story about prohibiting transport Story about destruction of of being involved in liquidations murdered at school of ammonia by train in certain unused domestic appliances areas

To make the 16 selected stories suitable the study by Kleemans, Van Cauwenberge, for the experiment, the spoken introduc- Hendriks Vettehen, and d’Haenens (2008). tions by anchorpersons were replaced by screenshots with written introductions. 3.2 Participants and procedure For each news story, this written intro- A total number of 288 higher educated duction summarized in one sentence the participants (48.6% male) took part in information that was originally given by the study. To represent young viewers, the anchorperson, as viewers needed a 96 university students were recruited (age short introduction to contextualize the range 18–26 years, M = 21.70; SD = 2.05). news story. We therewith prevented that To represent middle-aged viewers, 96 ac- the different news anchors would have af- ademic employees took part in the study fected the results. In addition, a few of the (30–50 years old, M = 41.91; SD = 4.85). The 16 stories were edited to enlarge the dif- oldest group (aged 60 years and above) ferences in packaging between the tabloid consisted of 96 former academic employ- and standard stories. To be more specific, ees or other higher educated elderly peo- we added or excluded some camera shots ple (M = 68.82; SD = 6.98). or included music or a sound effect, such The 16 news stories that were selected as the sound of a siren. After these minor for the experiment were divided into four changes, tabloid stories included a signifi- groups. Each group consisted of four sto- cantly larger number (t [7] = 10.73, p = .001) ries, one for each combination of content of camera shots per minute (M = 14.08; and packaging. Twenty-four participants SD = 1.26) compared to standard stories of each age category watched the stories (M = 7.79; SD = 0.56). Moreover, a combi- of one group. These stories were presented nation of music onsets, fades, wipes, and in a random order. The experiment was dissolves was present in the tabloid pack- conducted in an individual setting. Each aging stories, but not in the standard pro- participant first answered a few general duced stories. Although the original sto- questions about, for instance, age and sex, ries were edited, they still looked realistic. then watched the first story, and finally For example, previous research showed answered questions intended to measure that the mean number of camera shots in perceptions towards the story. This combi- news from the NOS Journaal contained nation of watching a story and answering 14.72 (SD = 8.66) camera shots per minute, some questions about it was repeated for implying that the camera changes for the the other three stories. standard and tabloid packaged stories in the current study felt within one standard deviation from the number reported in Kleemans et al. / Studies in Communication Sciences 17.2 (2017), pp. 183–194 189

3.3 Dependent variables 4 Results Participants were asked to rate seven statements about the perceived news me- Repeated measures MANOVAs were car- dia performance of the stories on a scale ried out for perceived news quality and ranging from 1 (totally disagree) to 7 (to- perceived sensationalism separately, with tally agree). Our selection of news media content (arousing vs. non-arousing) and performance indicators was first inspired packaging (tabloid vs. standard) as within- by previous studies on the topic: partici- subjects factors, and age (young, middle, pants indicated the extent to which a story old) as between-subjects factor. was informative, objective, and credible (cf. Grabe et al., 2000, 2003). Second, we 4.1 Effects of sensationalist news on included a statement that reflects the cen- perceived news performance tral concept as used by Wang and Cohen Hypothesis 1a, predicting that viewers (2009): perceived sensationalism. Third, perceived the performance of arousing we added two indicators that were derived content stories as more negative than from public discussions regarding the non-arousing content stories, was con- performance of arousing television news firmed by the results. Viewers perceived (cf. Bernstein, 1992; Fortuyn, 2000; Jones, the news quality of non-arousing sto- 2009): exaggerated and blown up. Finally, ries (M = 5.20; SE = 0.05) to be higher than we assessed the reliability of the story, be- the quality of arousing stories (M = 4.96; cause this issue reflects the background SE = 0.05), F (1.285) = 17.40; p = .001, r = .24. of the study, namely the decreased public In addition, arousing stories (M = 3.34, trust in today’s news, more directly. SE = 0.06) were perceived as considerably A principal component analysis with more sensational than non-arousing sto- oblique rotation of the seven statements ries (M = 2.54, SE = 0.05), F (1.285) = 140.41; was conducted for each type of news story p < .001, r = .57. separately. Both Kaiser’s criterion of retain- Hypothesis 1b was partially sup- ing factors with eigenvalues greater than 1 ported. No significant results of packag- and the criterion of factor loadings great- ing for the perceived quality of the news er than .60 (cf. Kline, 1994) implied that were found. However, tabloid packag- two components of perceived news media ing affected perceived sensationalism, performance could be discerned. The first, F (1.285) = 21.42; p < .001, r = .26. Findings which we labeled perceived news quality, indicated that tabloid produced stories consists of the statements regarding the (M = 3.10, SE = 0.06) were perceived as perceived informativeness, objectivity, more sensational than standard produced credibility, and reliability of the news story. stories (M = 2.78, SE = 0.06). The significant Cronbach’s alpha of the items varied be- interaction between content and pack- tween α = .82 and α = .84 for the four types aging, F (1.285) = 16.64; p < .001, further of news stories. The statements regarding refines this main effect. Post hoc F-tests the perceived sensational, exaggerated, showed that only for non-arousing stories, and blown up character of the news stories the inclusion of tabloid features increased together constitute the second dimension, perceived sensationalism (rM = 0.58; called perceived sensationalism. Cron- F (1.285) = 36.237; p < .001, r = .34). For bach’s alphas varied between α = .66 and arousing content, no significant (p = .577) α = .76 for the four types of news stories. increase in perceived sensationalism was For each type of news story, we calculated found due to packaging. the participant’s mean score to construct the variables perceived news quality and 4.2 The moderating role of age perceived sensationalism. Hypothesis 2a posed the expectation that older viewers had more negative perceptions regarding the media performance of arousing content stories than younger viewers. We found a main effect of age for 190 Kleemans et al. / Studies in Communication Sciences 17.2 (2017), pp. 183–194

Figure 1: Perceived news quality by content Figure 2: Perceived sensationalism by among young, middle-aged, and content among young, middle-aged, old viewers and old viewers

7 7

6 6

5 5.24 5 5.23 5.22 5.12 4.99 4.67 4 4

3 3 3.57 3.43 Perceived news quality

Perceived sensationalism 3.04 2 2 2.75 2.36 2.51 1 1 Old Middle-aged Young Old Middle-aged Young Viewers Viewers Arousing Non-arousing Arousing Non-arousing content content content content

perceived news quality, F (2.285) = 4.92; a lower news quality than non-arousing p = .008, which was further refined by the content stories, whereas older viewers interaction between content and age, perceived arousing content stories as hav- F (2.285) = 5.17; p = .006. As shown in Figure ing an equal news quality compared to 1 (means reported within bars, SE = 0.09 non-arousing content stories. Moreover, for all means), for older viewers there was older viewers were less critical regarding no difference (p = .968) in perceived news the sensational character of arousing con- quality between arousing and non-arous- tent news stories than middle-aged and ing stories. Middle-aged viewers perceived younger viewers. Hypothesis 2b also needs non-arousing stories as having somewhat to be rejected because no age differences more news quality than arousing stories, in viewer perceptions of both news quality F (1.285) = 20.91; p < .001, r = .26. The same and sensationalism with regard to packag- pattern was found for young viewers, ing were found. F (1.285) = 6.82; p < .001, r = .15. Regarding perceived sensationalism, a main effect for age was found, 5 Discussion F(2.285) = 7.05; p = .001. In addition, an interaction effect between content and age 5.1 Conclusion (F [2.285] = 3.63; p = .028) appeared. Post- This study investigated viewer evaluations hoc comparisons showed that participants of sensationalist television news. The most within all three age categories perceived important finding is that arousing content arousing content stories as more sensa- in news tends to be more negatively eval- tional than non-arousing content stories uated (in terms of both perceived news (F [1.285] = 32.88; p < .001, r = .32 for young quality and perceived sensationalism) viewers; F (1.285) = 81.75; p < .001, r = .47 for than non-arousing content. In addition, middle-aged viewers, and F (1.285) = 33.04; tabloid packaging increased perceptions p < .001, r = .32 old viewers). However, as of sensationalism, but only for non-arous- Figure 2 shows, middle-aged viewers ap- ing content stories. The relationship be- pear to be a bit more sensitive to arous- tween content of news stories and both ing news content than older and younger perceived news quality and perceived viewers (means reported within bars, SE sensationalism varied between viewers of varied between 0.09 and 0.11). different ages. Critical views on arousing In all, we need to reject H2a because content appeared to be particularly visible younger and middle-aged viewers per- among young and middle-aged viewers. ceived arousing content stories as having Kleemans et al. / Studies in Communication Sciences 17.2 (2017), pp. 183–194 191

5.2 Discussion Another limitation concerns the origin Some unexpected results provide a qual- of the stimuli. All stories were broadcast by ification of our overall conclusion. First, the same public service broadcaster. Since our finding that only tabloid produced we know from previous research that com- stories containing non-arousing content mercial broadcasters are more frequently were perceived as more sensational is a bit associated with the inclusion of arousing different from the Wang and Cohen (2009) content and tabloid features (Hendriks finding. Moreover, in contrast to Grabe and Vettehen et al., 2005), the results might be colleagues (2000, 2003), we did not find an somewhat more distinct when using news effect of packaging on the perceived qual- stories broadcast by commercial channels ity of news stories. This might possibly (cf. Leshner, Reeves, & Nass, 1998). Future be explained by the European context in research should, therefore, shed more which our study took place. In particular light on differences in viewer perceptions compared to US news stories, Dutch news of news performance between public ser- stories typically do not contain many tab- vice and commercial broadcast television loid features. The absence of effects here news. might thus be due to relatively smaller dif- Regardless of these limitations, this ferences in our stimulus materials. study contributes to an emerging body Second, the results regarding the of empirical evidence on perceptions of moderating role of age were different from sensationalist television news. Only a few what we expected. Not older adults, but studies had been conducted so far, and young and middle-aged viewers appeared each of these studies was somewhat lim- to be more critical on the use arousing con- ited. In the Wang and Cohen survey (2009), tent. An explanation may be found in the respondents were questioned about a fact that we selected stories from the main number of news categories (e.g., crime and public news program in The Netherlands. conflicts, accidents and disasters), without For over 60 years, this news program has actually being exposed to these news cate- been seen as a hallmark of high-quality TV gories. The Grabe et al. studies (2000, 2003) news. It could be that older viewers take included only a small number of partic- an acceptable news media performance of ipants and a small number of questions this particular news program for granted, to measure aspects of performance. The since they were raised at a time when the current study applied the rigorous exper- quality of the program was undisputed. imental methodology from the Grabe et Some critical remarks need to be al. studies, while using (a) a more compre- made. First, only higher educated viewers hensive set of evaluation questions that participated in our study. One might ask empirically allowed for distinguishing two whether this limits the degree to which performance dimensions, and (b) a larger our findings can be generalized. We know number of participants, making it possible from previous research that higher edu- to investigate the moderating role of age. cated viewers have generally more inter- To conclude, the overall study find- est in news than lower educated viewers ing that arousing television news con- (cf. Ettema & Kline, 1977; Liu & Eveland, tent and – to some extent – tabloid pack- 2005). Higher educated viewers are thus aged television news tends to be more more motivated to acquire information. negatively evaluated than non-arousing One might argue that they are hence more content and standard packaged news critical towards news performance (cf. contributes to the debate on the waning Bucy, 2003), which may have somewhat confidence in the news media by provid- limited the degree to which the results of ing an alternative to existent explana- the current study might be generalized. We tions of distrust. Until now, the broader recommend including participants with tendency to distrust institutions, experts, lower educational backgrounds in future and politicians (e.g., Bennett, et al., 1999; research as well. Jones, 2004), the tendency in political coverage to paint a cynical picture of politics 192 Kleemans et al. / Studies in Communication Sciences 17.2 (2017), pp. 183–194 by focusing on political scandals, politi- Broadcasting & Electronic Media, 53(2), cians’ private lives, and horse-race news 242–261. doi: 10.1080/08838150902907827. (Jones, 2005), and the possibility to con- Davis, H., & McLeod, S. L. (2003). Why humans sume online information that argues with value sensational news. An evolutionary and is critical of traditional news media perspective. Evolution and Human Behav- (e.g., Jones, 2005; Tsfati & Ariely, 2013) are ior, 24(3), 208–216. doi: 10.1016/S1090- discerned as factors explaining the declin- 5138(03)00012-6. ing trust in news media. The current study Emery, E. (1962). The press and America. An in- findings suggest that the increasingly sen- terpretative history of journalism (2nd ed.). sationalist character of television news Englewood Cliffs, N. J.: Prentice-Hall. might be an additional explanation of de- Ettema, J. S., & Kline, F. G. (1977). Deficits, creased perceptions of news performance. differences, and ceilings: Contingent con- ditions for understanding the knowledge gap. Communication Research, 4(2), 179– References 202. doi: 10.1177/009365027700400204. Eurobarometer (2016). Eurobarometer 86: Me- Adams, W. C. (1978). Local public affairs dia use in the European Union. Retrieved content on TV news. Journalism from http://ec.europa.eu/commfrontof- Quarterly, 55(4), 690–695. doi: fice/publicopinion/index.cfm/Survey/ 10.1177/107769907805500405. getSurveyDetail/search/media/sur- Baran, S. J., & Davis, D. K. (2012). Mass commu- veyKy/2137. nication theory. Foundations, ferment, and Fortuyn, W. (2000). De slag om het nieuws [The future (6th edition). Canada: Wadsworth battle for the news]. Den Haag: SDU. Cengage Learning. Grabe, M. E., Lang, A., & Zhao, X. (2003). News Baum, M. A. (2002). Sex, lies, and war: How content and form. Implications for mem- soft news brings foreign policy to the in- ory and audience evaluations. Commu- attentive public. The American Political nication Research, 30(4), 387–413. doi: Science Review, 96(1), 91–109. doi: 10.1017/ 10.1177/0093650203253368. S0003055402004252. Grabe, M. E., Zhou, S., & Barnett, B. (2001). Bek, M. G. (2004). Research note: Tabloidiza- Explicating sensationalism in television tion of news media. An analysis of tele- news: Content and the bells and whistles vision news in Turkey. European Journal of form. Journal of Broadcasting & Elec- of Communication, 19(3), 371–386. doi: tronic Media, 45(4), 635–655. doi: 10.1207/ 10.1177/0267323104045264. s15506878jobem4504_6. Bennett, S. E., Rhine, S. L., Flickinger R. S., & Grabe, M. E., Zhou, S., Lang, A., & Bolls, Bennett, L. L. M. (1999). Video malaise P. D. (2000). Packaging television news: revisited: Public trust in the media and The effects of tabloid on information government. The Harvard International processing and evaluative responses. Journal of Press/Politics, 4(4), 8–23. doi: Journal of Broadcasting & Electronic 10.1177/1081180X9900400402. Media, 44(4), 581–598. doi: 10.1207/ Bernstein, C. (1992). The idiot culture. Reflec- s15506878jobem4404_4. tions of post-Watergate journalism. New Graber, D. A. (1994). The infotainment quotient Republic, 22–28. in routine television news: A director’s per- Bucy, E. P. (2003). Media credibility recon- spective. Discourse & Society, 5(4), 483–508. sidered: Synergy effects between on-air doi: 10.1177/0957926594005004004. and online news. Journalism & Mass Habermas, J. (2006). Political communication Communication Quarterly, 80(2), 247–264. in media society: Does democracy still doi: 10.1177/107769900308000202. enjoy an epistemic dimension? The impact Cho, J., Shah, D. V., Nah, S., & Brossard, D. of normative theory on empirical research. (2009). “Split screens” and “spin rooms”: Communication theory, 16(4), 411–426. Debate modality, post-debate coverage, Doi: 10.1111/j.1468-2885.2006.00280.x. and the new videomalaise. Journal of Hendriks Vettehen, P. G. J., Beentjes, J. W. J., Nuijten, K., & Peeters, A. (2011). Arous- Kleemans et al. / Studies in Communication Sciences 17.2 (2017), pp. 183–194 193

ing news characteristics in Dutch tele French news stories]. Tijdschrift voor Com- vision news 1990–2004: An exploration municatiewetenschap, 36(4), 301–319. of competitive strategies. Mass Commu- Kleemans, M., Hendriks Vettehen, P. G. J., Eis- nication & Society, 14(1), 93–112. doi: inga, R., & Beentjes, J. W. J. (2014). Enjoy- 10.1080/15205431003615893. ment of arousing television news: The role Hendriks Vettehen, P. G. J., Nuijten, K., & of age and sensation seeking. Communica- Beentjes, J. W. J. (2005). News in an age of tions, The European Journal of Communi- competition: The case of sensationalism cation Research, 39(1), 89–99. doi: 10.1515/ in Dutch television news, 1995–2001. commun-2014-0006 Journal of Broadcasting & Electronic Kline, P. (1994). An easy guide to factor analysis. Media, 49(3), 282–295. doi: 10.1207/ London: Routledge. s15506878jobem4903_2. Krippendorff, K. (2004). Content analysis: An Hendriks Vettehen, P. G. J., Nuijten, K., & introduction to its methodology (2nd ed.). Peeters, A. (2008). Explaining effects of Thousand Oaks, CA: Sage. sensationalism on liking of television news Lang, A. (2000). The limited capacity model stories. Communication Research, 35(3), of mediated message processing. Jour- 319–338. doi: 10.1177/0093650208315960. nal of Communication, 50(1), 46–70. doi: Hollifield, C. A. (2006). News media perfor- 10.1111/j.1460-2466.2000.tb02833.x. mance in hypercompetitive markets: An Lang, A., Bolls, P., Potter, R. F. & Kawahara, K. extended model of effects. International (1999). The effects of production pacing Journal on Media Management, 8(2), 60– and arousing content on the information 69. doi: 10.1207/s14241250ijmm0802_2. processing of television messages. Journal Hvitfelt, H. (1994). The commercialisation of of Broadcasting & Electronic Media, 43(3), the evening news: Changes in narrativ 451–475. doi: 10.1080/08838159909364504 technique in Swedish TV news. Nordicom Lang, A., Shin, M., Bradley, S. D., Wang, Z., Lee, Review, 15(1), 33–41. S., & Potter, D. (2005). Wait! Don’t turn that Jones, A. S. (2009). Losing the news. The future dial! More excitement to come! The effects of the news that feeds democracy. New York: of story length and production pacing in Oxford University Press. local television news on channel changing Jones, D. A. (2005). Shooting the messenger: Ex- behavior and information processing in a plaining public discontent with the news free choice environment. Journal of Broad- media. Politics & Policy, 33(2), 242–262. casting & Electronic Media, 49(1), 3–22. doi: 10.1111/j.1747-1346.2005.tb00642.x. doi: 10.1207/s15506878jobem4901_2. Jones, D. A. (2004). Why Americans don’t Leshner, G., Reeves, B., & Nass, C. (1998). trust the media: A preliminary anal- Switching channels: The effects of tele ysis. The Harvard International Jour- vision channels on the mental representa- nal of Press/Politics, 9(2), 60–75. doi: tions of television news. Journal of Broad- 10.1177/1081180X04263461. casting & Electronic Media, 42, 21–33. doi: Kleemans, M., Hendriks Vettehen, P. G. J., 10.1080/08838159809364432. Beentjes, J. W. J., & Eisinga, R. (2012). The Liu, Y., & Eveland, W. P. Jr. (2005). Education, influence of age and gender on prefer- need for cognition, and campaign in- ences for negative content and tabloid terest as moderators of news effects on packaging in television news stories. Com- political knowledge: An analysis of the munication Research, 39(5), 679–697. doi: knowledge gap. Journalism & Mass Com- 10.1177/0093650211414559. munication Quarterly, 82(4), 910–929. doi: Kleemans, M., Van Cauwenberge, A., 10.1177/107769900508200410. d’Haenens, L., & Hendriks Vettehen, P. McQuail, D. (1992). Media performance. Mass (2008). Op zoek naar verklaringen voor communication and the public interest. sensatie in het nieuws. Een vergelijking London: Sage. tussen Nederlandse, Vlaamse, Waalse, Mutz, D. C., & Reeves, B. (2005). The new vid- en Franse televisiejournaals. [Explaining eomalaise: Effects of televised incivility sensationalism in television news. A com- on political trust. American Political Sci- parison of Dutch, Flemish, Walloon, and 194 Kleemans et al. / Studies in Communication Sciences 17.2 (2017), pp. 183–194

ence Review, 99(1), 1–15. doi: 10.1017/ Effects on credibility. In M. J. Metzger, A. J. S0003055405051452. Flanagin, The John D. and Catherine T. Newhagen, J. E., & Reeves, B. (1992). The eve- MacArthur Foundation Series on Digital ning’s bad news: Effects of compelling Media and Learning. Cambridge, MA: The negative television news images on mem- MIT Press, 73–100. ory. Journal of Communication, 42(2), Tapscott, D. (2009). Grown up digital. How the 25–41. doi: 10.1111/j.1460-2466.1992. net generation is changing your world. New tb00776.x. York: McGraw-Hill. Pew Research Center (2013). Amid criticism, Tsfati, Y., & Ariely, G. (2014). Individual support for media’s “watchdog” role and contextual correlates of trust in stands out. Retrieved from http://www. media across 44 countries. Commu- people-press.org/2013/08/08/amid-criti- nication Research, 41(6), 760–782. doi: cism-support-for-medias-watchdog-role- 10.1177/0093650213485972. stands-out/. Wang, T. L., & Cohen, A. A. (2009). Factors af- Robinson, M. J. (1976). Public affairs television fecting viewers´ perceptions of sensation- and the growth of political malaise: The alism in television news: a survey study in case of “The Selling of the Pentagon.” Taiwan. Issues & Studies, 45(2), 125–157. American Political Science Review, 70(2), Zhao, X., Liu, J. S., & Deng, K. (2012). Assump- 409–432. doi: 10.2307/1959647. tions behind inter-coder reliability indices. Schaefer, R. J., & Martinez III, T. J. (2009). Trends In C. T. Salmon (Ed.), Communication in network news editing strategies from Yearbook, Vol. 36. New York: Routledge (pp. 1969 through 2005. Journal of Broadcast- 419–480). ing & Electronic Media, 53(3), 347–364. doi: 10.1080/08838150903102600. Schatz, H., & Schulz, W. (1992). Qualität von fernsehprogrammen. Kriterien und meth- oden zur beurteilung von programmqual- ität im dualen fernsehsystem. Media Pers- pektiven, Vol. 1992(11), 690–712. Shoemaker, P. J. (1996). Hardwired for news: Using biological and cultural evolution to explain the surveillance function. Jour- nal of Communication, 46(3), 32–47. doi: 10.1111/j.1460-2466.1996.tb01487.x. Slattery, K. L., Doremus, M., & Marcus, L. (2001). Shifts in public affairs reporting on the network evening news: A move toward the sensational. Journal of Broadcasting & Electronic Media, 45, 290–302. doi: 10.1207/s15506878jobem4502_6. Slattery, K. L., & Hakanen, E. A. (1994). Sen- sationalism versus public affairs. Con- tent of local TV news; Pennsylvania revisited. Journal of Broadcasting & Electronic Media, 38(2), 205–216. doi: 10.1080/08838159409364257. Sociaal Cultureel Planbureau (2015). Media: Tijd in beeld [Media: Time overview]. Retrieved from: https://www.scp.nl/Publi- caties/Alle_publicaties/Publicaties_2015/ Media_tijd_ in_beeld Sundar, S. S. (2008). The MAIN Model: A heuris- tic approach to understanding technology