The Effect of House of Cards Exposure on Political Trust and Cynicism The Role of Prior Exposure in Narrative Persuasion
by Mornon Paul Student Number: 10832807
Master Thesis
Erasmus Mundus Program: Journalism, Media and Globalisation University of Amsterdam (UvA)
Supervisor: Wojcieszak Magdalena July 31, 2015
Abstract
This article analyses the effect of exposure to the fictional political series House of Cards through an experimental design with a sample of young Europeans (N=80). Drawing from cultivation theory and psychology research, I add to the Extended Elaboration
Likelihood Model (EELM) of narrative persuasion by taking into account prior exposure to the experimental stimulus (Slater & Rouner, 2002). I therefore develop a model of moderated mediation in which, familiarity with the stimulus moderates the mediated relationship
between exposure and story-consistent beliefs through narrative immersion. The findings suggested that political fiction did not affect viewers’ levels of political trust and cynicism. However, the intensity of prior exposure positively moderated the relationship between exposure and identification with the main character.
Keywords: House of Cards, Political dramas, Narrative Persuasion, Identification, Prior Exposure.
Introduction
Increasing levels of political distrust in Western democracies have been a matter of concern for both politicians and researchers (e.g. Agger, Goldstein & Pearl, 1961; Citrin, 1974). Some have hypothesized that media have their share of responsibility in this phenomenon by promoting a negative vision of the political world, emphasizing the flaws rather the successes of the political life or indulging in sensationalism (e.g. Boukes & Boomgaarden, 2014; Cappella & Jamieson, 1996). Accordingly, scholars have
experimentally showed that the nature of news could affect audience’s political beliefs such as level of cynicism toward politics and trust in institutions (e.g. Adriaansen, van Praag & de Vreese, 2010; Cappella & Jamieson, 1994; de Vreese, 2004; de Vreese, 2005; Elenbaas & de Vreese, 2008; Kleinnijenhuis, 2006).
The purpose of this research is to examine a different type of media content. I want to see whether watching fictional series about the functioning – or the malfunctioning – of the political system can trigger similar effects on their audience. Indeed, given the impressive wave of new productions that appeared over the past decades and their increasing popularity, this specific genre has gained an unprecedented significance (among others see The West Wing, 1999; 24, 2001; The Wire, 2002; The Good Wife, 2009; Borgen, 2010; Homeland, 2011; The Newsroom, 2012; Veep, 2012). In this study, I examine Netflix’s show House of Cards (2013). Beyond their popularity, fictional political dramas are interesting to look at because they provide a different viewing experience than traditional media. House of Cards is an emblematic example both in terms of content and consumption practices.
Firstly, contrary to news programs, the information conveyed by entertainment fiction is not constrained by accuracy or the need to respect reality. Political dramas often blur the border between reality and fiction. They display invented storylines but feature real
& Parris-Giles, 2002). Furthermore, the creators of these shows often highlight the quality and the verisimilitude of their production and claim that the events depicted are inspired from reality. This is for instance the case with House of Cards (e.g. Sneed, 2013). Therefore, it cannot be discarded that political series represent an indirect source of policy-relevant information for their viewers (Mutz & Nir, 2010).
Secondly, with the rise of convenience technologies, which empowered media consumers, exposure to entertainment programmes has deeply evolved (Lotz, 2007). Self-selected exposure has become an increasingly important topic of questioning for scholars (Bennet & Iyengar, 2008 Holbert, Garrett & Gleason, 2010). Henceforth, viewers have greater control on what they expose themselves to, thanks to the spread of new tools such as streaming and downloading services. Thereupon, Matrix (2014) developed the idea of a Netflix Effect from the name of the famous streaming platform. Viewers’ choose the extent and the intensity of the viewing experience. In the case of House of Cards, it resulted in increasing excessive practices, close to psychological addiction, often referred to as media binging (Devasagayam, 2014).
Consequently, it appears crucial to evaluate the extent to which political beliefs can rely on the viewing of fiction. In a nutshell, this research argues that the viewing of highly addictive and elaborated political series, featuring verisimilar situations, will undoubtedly influence viewers' image of real-world politics. I argue that the new generation of political dramas may have the persuasive power to trigger the endorsement of story-consistent beliefs. More precisely, in this study, I investigate the impact of exposure to political fictional shows from the standpoint of persuasion theory in order to analyse the way viewers process the information conveyed by fictional political series. Using an experimental design with a sample of N=80 young Europeans, I aim to answer the following research question:
RQ: How exposure to political fictional dramas affects viewers’ real-world political beliefs and to what extent prior-exposure moderates the endorsement of story-related beliefs?
My first goal is to confirm that political fiction has the persuasive potential to influence viewers’ vision of real-world politics. Many studies hypothesized and
experimentally found that exposure to fictional programs could alter viewers’ beliefs (e.g. Igartua & Barrios, 2012; Slater, Rouner & Long, 2006; Mutz & Nir, 2010). However, when specifically looking at political beliefs – cynicism in particular – scholars failed to
experimentally verify their predictions (e.g. Sigelman & Sigelman, 1974; Kaid, Towers & Myers, 1981). In this research, I draw from the Extended Elaboration Likelihood Model (EELM) of narrative persuasion to detect changes in viewers’ vision of the functioning of politics (Slater & Rouner, 2002). According to this model, I expect that exposure to political dramas will lead to the endorsement of story-consistent beliefs and that this relation will be partially mediated by viewers’ level of immersion in the fictional narrative (Cohen, 2001, Green & Brock, 2000, Slater & Rouner, 2002; Slater, Rouner & Long, 2006,). Story-consistent political beliefs will be evaluated by measuring viewers’ level of trust in institutions and cynicism towards politicians – outcomes that are relevant to the specific fictional drama studied. Indeed, House of Cards displays a very negative picture of American Politics.
My second goal is to add a component to the traditional model of persuasion. In this respect, this research aims to reduce a frequent methodological shortcoming of experimental designs looking at television series. I intend to evaluate the influence of prior-exposure to the experimental stimulus – as measured by participants’ self-reports – in the framework of the EELM. In order to bring this new dimension to the EELM, I bring inputs from cultivation theory as well as psychology studies looking at content familiarity and parasocial
way viewers process the information received during the experimental exposure (Bilandzic & Busselle, 2008 ; Cohen, 2001 ; Giles, 2002 ; Green et al., 2008 ; Hoffner & Buchanan, 2005 ; Slater & Rouner, 2002). In a theoretical section, I suggest a moderated mediation model in which prior exposure to political dramas positively moderates their persuasive potential. By doing so, I also try to better understand the reality of viewing practices.
Theory EELM: The Persuasive Power of Fiction
As Laer, de Ruyter, Visconti and Wetzels note, “nothing is less innocent than a story” (2014, p.798). Contrary to political speeches, editorials and other intentionally persuasive contents, fictions are not originally designed to persuade. Yet they do. Slater and Rouner (2002) developed the EELM to take into account the specificities of fiction as persuasive content. Their particularity resides in the use of narratives, which imply an altered and distinctive processing of the message received during exposure.
Classic theory of information processing like the Elaboration Likelihood Model (ELM) suggests that persuasive messages need to overcome existing cognitive barriers (Petty & Cacioppo, 1986). In The ELM, persuasion happens after either a careful cognitive
processing of the quality of the message (the central route) or can be affected by the persuasion context – e.g. the attractiveness of the source – (the peripheral route).
In the case of fiction, the differentiation between the central and the peripheral routes is not useful anymore. The acceptance of story-related beliefs depends on the absorption of the viewer in the narrative. Absorption mostly depends upon the objective quality of the narrative. If the audience member sees the fiction as a deliberate persuasive content poorly created, his or her cognitive barriers are reactivated (Slater & Rouner, 2002).
Narratives are persuasive because absorption alters the way viewer’s process information. Slater and Rouner (2002) explain that counter-arguing is fundamentally
incompatible with engagement in a story. In other words, when viewers are trapped in their favourite show, they are not able to intellectually reflect on the message they receive. Therefore they are more likely to assimilate this message and consequently to adopt story consistent-beliefs.
Political Dramas: A Blind Spot in the Field of Narrative Persuasion
Several studies have highlighted the potential for fiction to affect beliefs on a variety of topics such as immigration, religion, death penalty, sexual minorities or even the American justice system (e.g. Igartua & Barrios, 2012; Igartua, 2010; Green, 2004; Slater, Rouner & Long, 2006; Mutz & Nir, 2010).
Scholars often hypothesized that political beliefs such as trust or cynicism could be influence by the viewing of specific fictions looking in depth at the functioning of politics. Mulligan & Habel (2012) suggested that fiction could increase cynicism. Conversely, watching The West Wing was suggested as a possible remedy to help rejecting cynical depictions of politics by offering an engaging and idealistic vision of a fictional American President (Holbert et al., 2005; Parry-Giles & Parry-Giles, 2002).
However, when it comes to experimentally validating these hypotheses – notably increased levels political cynicism – scholars often failed to produce significant results (e.g. Kaid, Towers & Myers, 1981; Sigelman & Sigelman, 1974). Ultimately, to my knowledge, only one experimental study found clear causation between exposure to a show depicting fictional politics and change in political beliefs (Holbert et al., 2003). The authors showed that viewing The West Wing could indirectly play a role in improving trust in democratic institutions by priming a better image of the presidency.
In order to quantify the influence political dramas or movies, academics often tried to base their researches on surveys or to transpose theoretical frameworks looking at news effects such as agenda-setting, priming or framing (e.g. Moy and Pfau, 2000; Holbert et al.,
2003, Holbert et al., 2005, Holbrook & Hill, 2005). My goal in this study is to apply persuasion theory to the study of an overlooked political genre – namely political fiction. Furthermore, I intent to detect story-consistent beliefs that have been studied in depth by news effect researchers but remained disregarded by narrative persuasion theory – namely trust in institutions and political cynicism. Hence, I expect that the input of narrative persuasion will allow me to detect significant effects.
Story-Consistent Beliefs in House of Cards
In their model of narrative persuasion, Slater and Rouner (2002) argue that the suppression of counter-arguing caused by narrative immersion will lead to the development of story-consistent beliefs. They subsequently experimentally validated their hypothesis (Slater, Rouner & Long, 2006). In this study, I therefore expect that viewers exposed to the series House of Cards will adopt a vision of politics close to the one displayed in the show.
The Netflix’s drama gives a very particular picture of the underlying reality of American politics through the eyes of Frank Underwood, a fierce Congress Member. In House of Cards, the political game is a battlefield dominated by psychological – and physical – violence. Manipulations, revenge and elimination of rivals are commonly depicted practices in the show. In order to evaluate viewer’s vision of politics, I measure two commonly used indicators: trust in institutions and cynicism towards politicians. I also base my hypotheses on studies looking at political news. I argue that a ‘Frank Underwood Effect’ will grow among the viewers of House of Cards. Accordingly, the viewing of the show will increase their political cynicism and lower their trust in political institutions (see path1 red on Figure1).
Following Kleinnijenhuis (2006), “trust refers to a high estimation of the competence, honesty, or reliability of the one who is trusted, according to the expectations or norms of the beholder” (p.86). Hence, political trust in institutions implies an evaluation of how well institutions operate according to voters’ expectations (Hetherington, 1998). I claim that the
discourse developed in House of Cards is detrimental to trust. For instance, the show repeatedly borrows the language of war and competition. Politics are described as a race to acquire power and keep it by all means. Similar discourses found in news coverage, referred to as negative news and strategy coverage, have been found to decrease trust in politics (Cappella & Jamieson, 1994; Kleinnijenhuis, 2006). I expect that watching House of Cards will trigger similar effects. This leads to my first hypothesis:
Hypothesis 1a: Participants experimentally exposed to House of Cards will show lower levels of trust in their national institutions than an equivalent control group.
Political cynicism is defined as “strong distrust in the reliability and competence of politicians” (Adriaansen, van Praag & de Vreese, 2010, p.4). By reliability, the authors mean integrity, the extent to which politicians hold their promises and whether they act in the public interest. The competence of politicians is determined by their ability to do their job, the extent to which they take charge of problems and whether they know what is important for the people. There has been quite an extensive body of research on how strategic news, conflict-oriented frames or soft-news increase political cynicism (e.g. Boukes &
Boomgaarden, 2014; Cappella & Jamieson, 1996; de Vreese, 2004; Elenbaas & de Vreese, 2008). Political satire has also been criticized for promoting cynicism (Hart & Hartelius, 2007). Since House of Cards highlights the perfidy and the deceit of the political exchequer, I claim that viewing the show will increase cynicism towards politicians. Consequently, this leads to my second hypothesis:
Hypothesis 1b: Participants experimentally exposed to House of Cards will show higher levels of cynicism towards politicians than an equivalent control group.
Narrative Immersion: Transportation and Identification
As mentioned above, these persuasive effects are expected to emerge when viewers are truly immersed in the story. In other words, as the EELM explains, the persuasive process
is mediated by immersion in the narrative. Slater, Rouner and Long (2006) experimentally showed that the relation between narrative exposure and story-consistent beliefs was indeed at least partially mediated by immersion. Narrative immersion is divided in two mechanisms: transportation and identification, which have been theoretically and experimentally
differentiated (Tal-Or & Cohen, 2010). In this study, I aimed to confirm this mediation. If such mechanism was found significant, the results of the first set of hypotheses (H1a and H1b) should be ignored, as the indirect effect would replace the direct effect initially hypothesized.
Transportation refers to the process in which the viewer is entirely focused on the events occurring in the story rather than on the self (Green et al., 2008). When transported, the viewer loses contact with real world-facts, mentally and physically (Laer, de Ruyter, Visconti & Wetzels, 2014). Through this, fictional events might become more easily accessible cognitively for the viewer than reality (Green & Brock, 2000). Consequently, when viewers feel empathy for the characters because they imagine being part of the story-plot, their beliefs may be altered in a way consistent with the story (Laer, de Ruyter, Visconti & Wetzels, 2014). Accordingly, experimental designs showed that the more one was
transported in a narrative, the more he or she found the story realistic and developed story-consistent beliefs (e.g. Green, 2004; Green & Brock, 2000). Similarly, when looking specifically at television drama, transportation was found to be a good indicator to predict attitudinal and behavioural change (Murphy, Frank, Moran & Patnoe-Woodley, 2011, Slater, Rooner & Long, 2006). In accordance with transportation theory and these previous findings, I aim to confirm that transportation is the condition for developing story-consistent beliefs in the field of political fiction. Therefore, transportation would be a mediator of the relation between exposure and levels of trust and cynicism (see path2a blue on figure1). This leads to the following hypothesis:
Hypothesis 2a: The relationship between exposure and story-related beliefs will be mediated by transportation into the narrative.
Narratives also have the power to create vicarious social relationships through
identification with the fictional characters of the story. While transportation could be defined as loss of real-life awareness, identification is tantamount to loss of self-awareness.
According to Cohen (2001), “identification is a process that culminates in a cognitive and emotional state in which the audience member is aware not of himself or herself as an audience member, but rather imagines being one of the characters” (p.252). Hence, the viewer’s initial beliefs might be altered since he or she is likely to share the feelings, the perspective and the goals of the character he or she identifies with. Accordingly,
experimental settings highlighted that attitudinal shift was correlated with the level of identification with the main protagonist (Igartua, 2010; Igartua and Barrios, 2012). It was also previously experimentally highlighted that identification could mediate the relationship between exposure to fiction and story-related beliefs (Chung & Slater, 2013; de Graaf, Hoeken and Beentjes, 2011; Slater, Rouner and Long, 2006). I intend to confirm this mediated mechanism in the field of political fiction (see path2b blue on figure1). This leads to the following hypothesis:
Hypothesis 2b: The relationship between exposure and story-related beliefs will be mediated by identification with the main character.
The Flaws of Persuasion Theory in the Analysis of Repeated Exposure
Long-term exposure to a specific program is clearly a blind spot of persuasion theory. Slater and Rouner (2002) do not explicitly include this component in the EELM. They only hypothesize that the stronger the stimulus, the stronger the effect. Consequently, my second and overarching goal is to add to the EELM by looking at the influence of prior exposure on the persuasive process. Studies evaluating the impact of a single episode ignore the modality
of real-life exposure and therefore display a very low ecological validity. In order to add to the EELM model, I look at other theoretical inputs such as cultivation theory and psychology studies.
Looking at long-term effects of television viewing, cultivation research has shown that repeated exposure might have important consequences on viewers’ vision of the real world (see Morgan & Shanahan, 2010 for a literature review). For instance, a relationship was found between viewing local television and fear of crime (Romer, Jamieson & Aday, 2003). Similarly, repetitive exposure to fictional programs featuring paranormal events was shown to be a significant predictor of paranormal beliefs (Sparks, Nelson & Campbell, 1997).
In psychological studies, familiarity with a specific message was shown to alter the way we process this message. For instance, Zajonc (1968) argued that repeated exposure to the most basic stimuli, such as words, is a sufficient condition to enhance attitudes toward this message. Novelty is inherently related to uncertainty and preconceived negative effects. However, when a stimulus becomes familiar, it also becomes more attractive. Marketing studies similarly found that message repetition could result in more positive attitudes and higher ad-related beliefs (e.g. Hawkins, Hoch & Meyers-Levy, 2001, McCullough & Ostrom, 1974).
More elaborated messages such as complex narratives have been hypothesized to trigger similar effects. Schiappa, Gregg and Hewes (2006) showed a positive correlation between the viewing frequency of a show giving a positive picture of the gay community and low levels of prejudice toward gay men. Bilandzic and Busselle (2008) argued that the strong stimulus created by repeated exposure to certain fiction genres might lead to the development of stronger story-consistent beliefs rather than that of a single exposure. According to them, the more viewers are familiar with the main message developed by a program, the less cognitively critical they should be toward it.
Accordingly, I argue that the more one watches House of Cards, the less he or she will be able to counter-argue with the message displayed by the show. The strength of the stimulus can be divided in two distinct notions. Firstly, the extent of exposure in this context relates to the number of episodes of the series the viewer has previously watched. Secondly, the intensity of exposure relates to the frequency or the number of episodes the viewer usually watches per week – or used to watch when he or she was following the show. I hypothesize that both the intensity and the extent of exposure will positively moderate the endorsement of story-related beliefs (see path3 green on Figure1).
Hypothesis 3a: Participants who report high levels of prior extensive exposure to House of Cards will show higher levels of story-consistent beliefs than viewers who report low levels of prior exposure or no prior exposure.
Hypothesis 3b: Participants who report high levels of prior intensive exposure to House of Cards will show higher levels of story-consistent beliefs than viewers who report low levels of prior exposure or no prior exposure.
The Effect of Prior-Exposure on Transportation and Identification
My third goal is to take a closer look at the influence of prior exposure on the mediation process described above. I hypothesize that both transportation and identification will be positively affected by prior exposure to the stimulus material.
Firstly, the level of transportation may be a function of the familiarity with the content. Bilandzic and Busselle (2008) argued that since transportation is an enjoyable experience, viewers are likely to repeatedly choose the very same topic or genre that
transported them. Repeated exposure would increase knowledge and interest in the field and refine familiarity with genre conventions – e.g. political fiction. Thereafter, it would facilitate the processing of the story and consequently increase transportation and so forth. In a second experiment by Green et al. (2008), it was shown that viewers who were exposed twice to the
same storyline were more transported. Participants exposed to a novel and then to the film version of the same story were more transported than the non-readers. In this study, I want to test whether repeated exposure to the same medium triggers similar effects. In accordance with these previous works, I expect that intense and extensive previous exposure will positively moderate the level of viewers’ transportation (see path4 green on figure1). Hypothesis 4a: Participants who report high levels of prior extensive exposure to House of Cards will show higher levels of transportation into the narrative than viewers who report low levels of prior exposure or no prior exposure.
Hypothesis 4b: Participants who report high levels of prior intensive exposure to House of Cards will show higher levels of transportation into the narrative than viewers who report low levels of prior exposure or no prior exposure.
Secondly, identification with characters is also likely to increase over time.
Psychology research looking at person perception has shown for a long time that the greater attraction caused by familiarity led to higher perceived similarity and likeableness (Moreland & Zajonc, 1982). Similar mechanisms are hypothesized for invented characters. Research looking at binge watching practices argued that regular viewers were more likely to develop high levels of attraction and even addiction to their favourite characters (Devasagayam, 2014).
In the EELM, Slater and Rouner (2002) mentioned that certain formats such as long-running serial featuring extended narratives are likely to increase investment with characters. Similarly, in his model of parasocial interaction, Gilles (2002) explained that the relationship one can develop with a character evolves with successive encounters – typically the different episodes of a drama. During the viewing of one episode, the viewer may identify with a character, especially if he or she starts liking the character (Chory, 2013; Tian & Hoffner, 2010). When the viewer experiences several encounters, these attitudes are reinforced.
Typically a strong relationship will only occur after numerous ‘meetings’ with the characters (Gilles, 2002). Beyond a simple relationship, Cohen (2001) states that familiarity with a character should increase the likeliness of imagining being that character. Consequently, I expect that previous exposure will positively moderate the level of viewers’ identification with the main protagonist (see path5 green on figure1).
Hypothesis 5a: Participants who report high levels of prior extensive exposure to House of Cards will show higher levels of identification with the main character than viewers who report low levels of prior exposure or no prior exposure.
Hypothesis 5b: Participants who report higher levels of prior intensive exposure to House of Cards will show higher levels of identification with the main character than viewers who report low levels of prior exposure or no prior exposure.
Finally, in order to link the previous hypotheses, I develop a model of moderated mediation (see figure 1). I previously hypothesize that prior exposure will positively moderate transportation and identification (hypotheses 3 and 4). Therefore the following model is expected to emerge at the persuasive level (see figure6):
Hypothesis 6: The mediated relationship between exposure and story-consistent beliefs through narrative immersion will be moderated by prior exposure, such that participants who report high prior exposure will show higher levels of immersion and consequently higher levels of story-consistent beliefs.
[Figure 1 here]
Method Context
The current political context adds a particular resonance to this study. Indeed, in the aftermath of the economic crisis, political beliefs have been deeply altered. Trust in national institutions declined in most European countries (European Social Survey, 2013). Only 29%
of Europeans trust their government and 30% their national Parliament (Eurobarometer, Autumn 2014). The European youth is particularly touched by the phenomenon. As a matter of fact, last year, levels of cynicism toward politicians and politics reached 69.2 on a 0-100 scale while participation in politics only reach 18.3 (Myplace Thematic Report, 2015). Therefore, the findings of this study could increase the field of investigation looking at the possible influence of media in this concerning phenomenon.
Design and Procedure
This study relies on a pretest-posttest design with an experimental group and a control group, which guarantees strong internal validity. The experimental group viewed the first episode of House of Cards entitled ‘Chapter 1’ (Willimon, 2013). The control group viewed the first episode of Mad Men entitled ‘Smoke Gets in Your Eyes’ (Weiner, 2007), which was considered unlikely to affect the dependent variables – namely trust and cynicism. Indeed, in the chosen episode, there is neither characters playing politicians nor particular mention of the political life. Mad Men displays the private and professional life of Don Draper, the creative director of an advertising company.
The data were collected in several waves between the 19th of May and the 15th of June 2015. After completing the pretest questionnaire, participants were randomly assigned to the experimental or the control condition. One day after the pretest, participants were asked to watch the episode they were randomly assigned to. Right after the viewing, they completed the second questionnaire.
The data were collected in two ways. First, viewing sessions were organized to collect part of the responses (35% of the sample) with total control over the procedure. The show was screen at my home, in a living room. The participants were gathered in small groups (between 5 and 9). Drinks and food were provided to the participants. The rest of the data was collected online (65% of the sample). The episode and the posttest questionnaire were
only sent to the participants after they filled in the pretest. These participants watched the show individually, at their place. No significant mean differences were reported in any of the dependent or independent variables for these two exposure conditions. The mode of exposure did not alter the results of the experiment. Out of the 96 people invited, 80 completed the entire procedure (83.3%).
Questionnaires
In both groups, all participants completed similar pretest and posttest questionnaires. The pretest questionnaire included questions tapping the baseline level of the dependent variables (trust and cynicism) so that to assess the successful randomization of the sample distribution, as well as basic demographics (age, sex, education, nationality, political interest and political leaning). The posttest questionnaire measured transportation and identification as well as the dependent variables (trust and cynicism).
The moderator, prior exposure, was also measured at the posttest. This could
represent a substantial limitation of this study. Moderators are usually evaluated in the pretest questionnaire, as they should not be influenced by exposure. However, the particularities of this moderator make a case for this choice. On the one hand, the risk of sensitizing the participants by including this measurement in the pretest was high. Indeed, unveiling the nature of the stimulus and the overarching goal of the study one day before the actual exposure would have represented an important flow likely to influence the dependent variables and the mediators. On the other hand, the measurement of self-reported prior exposure was unlikely to be affected by exposure to the stimulus as they represented
quantified numbers, impervious to social desirability bias, rather than the personal evaluation of an attitude or a perception. Moreover, in order to improve answers’ accuracy, a note was provided to the participant in the questionnaire featuring the total number of episodes and the number of episode per season. Overall, the risk of sensitization was considered higher than
the risk of altering the moderator measurement. A MANOVA was performed to make sure that no difference could be found between the experimental and the control group. The test showed that there was no relationship between the condition and neither the self-reported
intensity of exposure (F(1, 80)=.33, p>.05, ηp2=.00), nor the self reported extent of exposure
(F(1, 80)=1.24, p>.05, ηp2=.02), It confirms that they could be tested as valid moderators.
Participants
Participants were recruited through convenience sampling and snowball sampling. They were indeed mostly personal acquaintances, fellow students or “friends of a friend”. The study used a sample of N=80 young Europeans between 17 and 34 years old, with an average age of 24 years old. There was roughly the same number of men and women (57.50% of women and 42.5% of men). A very large majority of the participants were bachelor or master students (86.3%). On a scale from 1 (left) to 7 (right), 48.75% declared themselves left wing, 13.75% right wing and 37.50% did not have a strong affiliation. When it comes to interest in news and public affair, assessed on a scale from 1 (not at all interested) to 7 (very interested), 83.75% of the participants declared themselves interested. A large proportion of the participants already knew House of Cards: 53.75% viewed at least one episode of the show before the experimental stimulus. The well educated and interested in politics are clearly over-represented in the sample. Moreover, even if 15 countries were represented in total, France (33.75%) and Spain (32.50%) were largely over-represented. Stimulus
House of Cards is an American political drama comprised of 3 seasons for a total of 39 episodes at the time of the experiment. The first episode ‘Chapter One’ displays the initial situation and the future plots of the show. It features Frank Underwood, the House Majority Whip, as main protagonist. At the beginning of the episode, he has ensured the election of the candidate of his party as new President of the United States. As a reward, he was promised
the position of Secretary of State but the new administration and the freshly elected president do not honour their agreement. Raging mad at the betrayal, Underwood crafts a plan to destroy its former allies and climb up the political ladder.
Measurements
Story-consistent beliefs.
This study adapted an existing 7-points scale of political trust (Mishler & Rose 2001). Respondents indicated from 1 (not at all) to 7 (a lot) how much they trusted different
institutions, namely political parties, government, parliament and the head of the executive. A principle component analysis showed that the four items formed a single uni-dimensional scale for the pretest (Eigenvalue=2.93, 73.25% of the variance explained) and the posttest (Eigenvalue=3.13, 78.23% of the variance explained). All items correlated with the first component. The higher the score, the higher was the trust in the regarded institution (pretest M=2.92, SD=1.23; postest M=3.00, SD=1.30). The reliability test indicated good consistency for the pretest with Cronbach’s α=.87 and excellent internal consistency for the posttest measurement with α=.90.
The cynicism scale developed Adriaansen, van Praag and de Vreese (2010) was adapted for this study. Participants were asked on 7-points scales, from 1 (strongly disagree) to 7 (strongly agree), whether they agree or disagree with diverse statements. Four items were extracted from the scale, such as “politicians consciously promise more than they can
deliver”. A principle component analysis showed that the four items formed a single uni-dimensional scale for the pretest (Eigenvalue=1.96, 48.93% of the variance explained) and the posttest (Eigenvalue=2.09, 52.17% of the variance explained). All items correlated with the first component. The higher the score, the higher was the level of cynicism towards politicians (pretest M=5.55, SD=.85; postest M=5.40, SD=.90). The reliability test indicated
acceptable consistency for the pretest with Cronbach’s α=.64 and the posttest measurement with α=.66.
Transportation.
Transportation in the narrative was measured with a scale developed by Green and Brock (2000). It consisted of a set of indicators measuring the extent of the absorption in the narrative. Participants were asked on 7-points scales, from 1 (strongly disagree) to 7 (strongly agree), whether they agree or disagree with statements related to their viewing experience. Six items were extracted from the scale, such as “I was mentally involved in the show while watching it”. A principle component analysis showed that the six items formed a single uni-dimensional scale (Eigenvalue=2.60, 43.25% of the variance explained). All items correlated with the first component. The higher the score, the higher was the transportation into the narrative (M=4.18, SD=1.05) The reliability test indicated good consistency with Cronbach’s α=.73.
Identification.
Identification with the main character was measured using Cohen’s (2001) scale. The questions were adapted to the main characters of House of Cards and Mad Men. Participants were asked on 7-points scales, from 1 (strongly disagree) to 7 (strongly agree), whether they agreed or disagreed with several statements. A principle component analysis showed that among the ten extracted items, two distinct uni-dimensional scales stood out (first
component: Eigenvalue=4.81, 48.10% of the variance explained; second component: Eigenvalue=1.26, 12.58% of the variance explained). Eight items such as “I think I have a good understanding of Frank Underwood” were finally used as the identification scale. The higher the score, the higher was the identification with the main character (M=3.93, SD=1.17) The reliability test indicated good consistency with Cronbach’s α=.87. The two other items were related to empathy. They were discarded from the analysis.
Previous exposure.
Two indicators were measured in order to evaluate the viewing history of participants. The extent of exposure measured the number of episodes of House of Cards the participant has watched before the experimental exposure (N=80, min=0, Max=39, M=11.53, SD=15.29). The intensity of exposure measures how many episodes of House of Cards the participant was watching per week during the period of exposure (N=80, min=0, Max=20, M=3.00, SD=4.29). Both extent and intensity were binary recoded by splitting the continuous variable at the mean score (1=high, 0=low).
The pretest questionnaire also contained questions about age, sex, education,
nationality and interest in news and public affairs. The exact wording of all items as well as the mean score can be found in the appendix (Table1).
Results
Before analysing the results, a series of Analysis of Variance (ANOVA) was performed to make sure that the randomization of the participants was effective in terms of demographics and initial levels of story-consistent beliefs. The tests showed that there was no relationship between the condition (experimental or control group) and the initial level of trust (F(1, 80)=.04, p>.05, ηp2=.00) or the initial level of cynicism (F(1, 80)=.97, p>.05, ηp2=.01).
The difference in trust and cynicism between conditions after exposure was therefore estimated by looking at the levels of these variables in the posttest. As far as the
demographics were concerned, the results showed that there was no significant difference between conditions for sex (F(1, 80)=3.33, p>.05, ηp2=.04), nationality (F(1, 80)=1.52, p>.05,
ηp2=.02), interest in politics (F(1, 80)=.09, p>.05, ηp2=.00) and ideology (F(1, 80)=1.96, p>.05,
ηp2=.02). The randomization was therefore successful.
Analysis
of exposure to House of Cards on story-related beliefs (H1a and H1b) and to see whether the endorsement of these beliefs was mediated by transportation (H2a) and identification (H2b). H1a and H1b were tested by between-subjects Analysis of Variance (ANOVA) in order to determine a statistical significance difference between exposure and endorsement of story-consistent beliefs. Thereafter, the mediation analysis (H2a and H2b) was conducted via PROCESS, a plugin macro for SPSS (Model 4: Hayes, 2013), to see if mediation occurred through transportation and identification between the predictor exposure and story-consistent beliefs. The indirect effects were estimated via bootstrapping 1000, by reporting the
bootstrapped confidence intervals (CI=95%).
My second goal was to evaluate the influence of previous exposure on the persuasive process. I predicted that the extent and the intensity of prior exposure would positively moderate the endorsement of story-consistent beliefs (H3a and H3b), the levels of
transportation (H4a and H4b) and identification (H5a and H5b). The moderation analyses were conducted by between-subjects Analysis of Variance (ANOVA) looking at the interaction between the variables and the two components of prior-exposure.
My third and last goal was to test whether the extent of prior exposure and the intensity of prior exposure could moderate the mediation model, in which transportation and identification mediate the effect of exposure on story-consistent beliefs. The moderated mediation analysis (H6) was also conducted via PROCESS (Model 9; Hayes, 2013). Hypothesis Testing
Hypothesis 1: Story-consistent beliefs.
The first set of hypotheses predicted that participants exposed to House of Cards would show higher levels of cynicism and lower levels of trust than participants exposed to Mad Men. Significant effects were found neither for trust in institutions (F(1, 80)=.04, p>.05
and H1b cannot be supported.
Hypothesis 2: Mediation through transportation and identification.
H2a and H2b predicted that the relation between exposure and trust or cynicism would be mediated by transportation and identification (Figure2). There was no significant indirect effect of exposure on trust through transportation (b=.-00, BCa CI [-.12, .08]) or through identification (b=.15, BCa CI [-.13, .22]) (see Figure2a). Similarly, there was no significant indirect effect of exposure on cynicism through transportation (b=.00, BCa CI [-.10, .15]), or through identification (b=-.08, BCa CI [-.29, .01]) (see Figure2b). Therefore H2a and H2b cannot be supported.
[Figure 2 here]
Hypothesis 3: Prior exposure and story-consistent beliefs.
With H3a, H3b, I hypothesized that the extent of prior exposure and the intensity of prior exposure would positively moderate the relationship between exposure and story-related beliefs. The interaction effect between the extent of prior exposure and experimental exposure was significant neither for trust (F(1, 80)=.24, p>.05, ηp2=.00), nor for cynicism (F(1, 80)=.59, p>.05, ηp2=.00). Similarly, the interaction effect between the intensity of prior
exposure and story related beliefs was significant neither for trust (F(1, 80)=.50, p>.05,
ηp2=.00), nor for cynicism (F(1, 80)=.06, p>.05, ηp2=.00). Therefore hypothesis H3a and H3b
cannot not be supported.
Hypothesis 4 and 5: Prior exposure and transportation/identification.
H4a, H4b and H5a, H5b tested whether the extent of prior exposure and the intensity of prior exposure to House of Cards moderated the relation between experimental exposure and transportation or identification. The interaction effect between extent of prior exposure and experimental exposure was significant neither for transportation (F(1, 80)=2.12, p>.05
intensity of prior exposure and experimental exposure was only approaching significance for transportation (F(1, 80)=3.64, p=.06 ηp2=.05). Therefore H4a, H4b, and H5a cannot be
supported. However, the interaction between intensity of prior exposure and experimental exposure was significant for identification (F(1, 80)=4.86, p=.03 ηp2=.06). Accordingly, in the
experimental group, participant reporting high prior intense exposure to House of Cards showed significantly higher levels of identification (prior intense exposure: M=4.81, SD=1.16, vs. no prior intense exposure: M=3.66, SD=1.27) (Figure 3).
[Figure 3 here] Hypothesis 6: Mediated moderation.
The analysis tested the mediated relationship between exposure and story-consistent beliefs through transportation and identification at different levels of the moderators. I
therefore obtained a series of four analyses (see Figure4 for an example). No significance was found when looking at the relationship between exposure and trust through transportation (in the condition high prior intensity of exposure: b=-.07, Boot SE=.14, Boot CI [.-.47, .13]; in the condition high prior extent of exposure b=.05, Boot SE=.13, Boot CI [-.14, .42]) or through identification (in the condition high prior intensity of exposure: b=.05, Boot SE=.25, Boot CI [.-.45, .61]; in the condition high prior extent of exposure b=-.01, Boot SE=.09, Boot CI [-.33, .11]. As for the relationship between exposure and cynicism, significance was found neither through transportation (in the condition high prior intensity of exposure: b=.17, Boot SE=.15, Boot CI [-.03, .61]; in the condition high prior extent of exposure b=-.14, Boot SE=.14, Boot CI [-.45, .08]), nor though identification (in the condition high prior intensity of exposure: b=-.24, Boot SE=.18, Boot CI [.-.69, .02]; in the condition high prior extent of exposure b=.06, Boot SE=.10, Boot CI [-.04, .40]). Consequently the moderated mediation model was rejected and H6 cannot be supported.
Discussion
This paper had two overarching goals. On the one hand, I aimed to see whether House of Cards could influence viewers’ levels of trust in institutions and cynicism towards
politicians. Persuasion theory – more precisely the Elaboration Likelihood Model developed by Slater and Rouner (2002) – was used to test whether the viewing of House of Cards could lead to the endorsement of story-consistent beliefs. On the other hand, I made an attempt to bring an original theoretical contribution. I wanted to estimate the effect of prior exposure on the persuasive process described in the EELM. Basing my hypotheses on psychological studies and the few innovative researches that tried to evaluate the influence of repeated exposure in the framework of narrative persuasion, I developed a model of moderated mediation, in which the extent and the intensity of prior exposure moderate the endorsement of story-related beliefs. By doing so, I intended to see whether extensive and intense
consumption practices influence the way one processes narrative messages.
None of the overarching hypotheses were supported. There was no direct effect between exposure to the show and endorsement of story-related beliefs (H1). In this study, watching House of Cards did not alter viewers’ levels of cynicism toward politicians or trust in institutions. Moreover, contrary to the prediction of the EELM, neither transportation, nor identification was found a significant mediator of the relationship between exposure to House of Cards and related political beliefs (H2). Besides, contrary to my expectations prior
exposure was not found a significant moderator of the direct relationship between exposure and story-consistent beliefs (H3). Eventually, prior exposure did not moderate either the indirect relationship between exposure and story-consistent beliefs through identification and transportation (H6).
Several factors could explain these results. As supported by Feldman and Sigelman (1985), “Some beliefs about the world may be quite immune to change because they are
central in an individual’s belief system or linked closely to other beliefs” (p.558). Political trust and cynicism represent deeply routed beliefs, maybe unlikely to be altered by sporadic exposure. They are social constructs influenced by a very large number of factors such as political interests, interpersonal trust, partisanship, education and more generally individuals’ social background (Agger, Goldstein & Pearl, 1961, Hetherington, 1998). In some way, the lack of effect is “reassuring” as the viewing of a fictional drama did not influence educated and politically sophisticated individuals – namely the very large majority of the sample.
Previously, mostly fictions with a clear ‘message’ or about contentious issues – such as religion, minorities, death penalty or immigration – were able to trigger significant changes in beliefs (e.g. Igartua & Barrios, 2012; Igartua, 2010; Green, 2004; Slater, Rouner & Long, 2006). Cynicism and trust might not represent the best indicators to look at because they are very general and vague concepts not only relevant to House of Cards. For instance Holbert et al. (2003) were able to detect significant political effects when specifically looking at the perception of the U.S. presidency in The West Wing.
Eventually, only one hypothesis was supported. Prior intense exposure positively moderated the relationship between exposure and identification – so much that viewers reporting prior intense exposure to House of Cards showed significantly higher levels of identification with the main character. This goes in line with our rationale predicting that identification would increase with content familiarity.
Although they are very limited, these findings have several implications.
Methodologically, they confirm that research looking at mainstream media content should not ignore prior exposure in experimental settings. Participants previously exposed should be either systematically discarded as in Igartua and Barrios’s study (2014) or prior exposure should be included as a valid moderator in the same way as gender, age or ideology.
Theoretically, this research gives a long-term relevance to the study of identification. Two studies by Bilandzic and Busselle (2008) and Green et al. (2008) have looked at
transportation in a long-term perspective, but identification remained overlooked. As it was highlighted for transportation, the findings of this study indicate that repeated exposure would enhance identification with the characters.
Finally, the results contribute to illustrate the crucial need to take new modes of media consumption into account. Nielsen reported that 88% of Netflix subscribers claimed watching regularly at least three episodes of a show per day (2013). In the sample of this study, the regular viewers of House of Cards declared watching more than five episodes per week when the show is released (M=5.71). Accordingly, this research points out that heavy viewers were shown to react differently to the experimental exposure as they felt more identify with the main protagonist of the show. Studies looking at television series need to acknowledge this revolution and elaborate experimental designs accordingly.
The limitations of this study are numerous and they might explain for a great deal the lack of significant results for almost all hypotheses. First of all, the sample might have been too small to estimate complex mechanisms such as moderated mediation. Convenience sampling caused the sample to mostly consist of Spaniard and French persons, highly interested in politics, well educated and left wing. It is therefore irrelevant to extrapolate the findings of the study beyond this population.
Secondly, the experimental design is not without flaws. A pretest-posttest design carries an inherent risk of sensitization as the participants are asked twice the same set of questions. Furthermore, since the posttest questionnaire was filled right after exposure, even if results would have been found, no long-term effects conclusions could have been made.
Thirdly, the measurement of the moderators constitutes a major limitation. Prior exposure was self-evaluated rather than experimentally manipulated. Consequently, the
direction of the causation between exposure and identification cannot be substantiated. Not only higher prior exposure would lead to higher identification but also higher identification might lead to higher levels of exposure.
Eventually, the very choice of House of Cards is problematic. This show was chosen for its popularity. Indeed, in order to evaluate the influence of prior exposure, I had to focus on a mainstream show with a broad audience. Yet, this choice led to one last important limitation: House of Cards depicts American politics while the population I was interested in is only European. Unfortunately, no European drama can claim to have the same broad viewership as House of Cards. I might have wrongly hypothesized that the chosen population would make analogies between national European politics and the power game as it is
described in House of Cards.
Future research should attempt to solve these limitations. First, trust and cynicism should probably be discarded as indicators of story-consistent beliefs in the experimental study of political fiction. In the very case of House of Cards, one could more closely look at the depiction of power displayed by the serial both in the private and public life of the main characters (Sternbergh, 2014). To detect effects, alternative theoretical framework might be of help such as priming, framing or even exemplification theory (Holbert et al., 2003, Holbert et al., 2005; Wojcieszak and Mutz, 2014). Finally, the consequences of prior exposure would be better investigated through longitudinal studies manipulating the intensity and the extent of exposure with a stronger control over the experimental procedure. In general, the study of media effects would gain relevance if it keeps using the input of different theories as outlined by Busselle and Bilandzic (2008) with cultivation and narrative transportation. In this
manner, I tried to bring interesting inputs from psychological study, news effect theory and cultivation to the study of narrative persuasion.
Conclusion
This research was motivated in the first place by a real debate started by various commentators and politicians. Although many critics have praised House of Cards for the quality of its storytelling, the show has also been accused of giving an overly simplified and distorted picture of American politics – “A vision of the American government, not as we wish it were, but as we secretly fear it is” (Sternbergh, 2014, antepenultimate paragraph). Some claimed that House of Cards would be the worst show to understand the real functioning of American politics or hypothesized that it could fuel already high levels of distrust and cynicism (Frank, 2013, Masket, 2015). Despite this questioning, no study had previously looked at the influence of this widely popular show. To this day, watching House of Cards was not proven of any influence on viewer’s political trust or cynicism and intensive prior exposure did not lead to significant results either. There is therefore no empirical
References
Adriaansen, M., van Praag, P., & de Vreese, C. (2010). Substance Matters: How News Content can Reduce Political Cynicism. International Journal Of Public Opinion Research, 22(4), 433-457. doi:10.1093/ijpor/edq033
Agger, R., Goldstein, M., & Pearl, S. (1961). Political Cynicism: Measurement and Meaning. J Of Pol,23(03), 477. doi:10.2307/2127102
Beavers, S. (2002). The West Wing as a Pedagogical Tool. PS: Political Science & Politics, 35(02). doi:10.1017/s1049096502000537
Bennett, W., & Iyengar, S. (2010). The Shifting Foundations of Political Communication: Responding to a Defense of the Media Effects Paradigm. Journal Of
Communication, 60(1), 35-39. doi:10.1111/j.1460-2466.2009.01471.x
Bilandzic, H., & Busselle, R. (2008). Transportation and Transportability in the Cultivation of Genre-Consistent Attitudes and Estimates. Journal Of Communication, 58(3), 508-529. doi:10.1111/j.1460-2466.2008.00397.x
Boukes, M., & Boomgaarden, H. (2014). Soft News With Hard Consequences? Introducing a Nuanced Measure of Soft Versus Hard News Exposure and Its Relationship With
Political Cynicism. Communication Research, 42(5), 701-731. doi:10.1177/0093650214537520
Cappella, J., & Jamieson, K. (1994). Public Cynicism and News Coverage in Campaigns and Policy Debates: Three Field Experiments. In American Political Science Association Conference. New York, NY.
Cappella, J., & Jamieson, K. (1996). News Frames, Political Cynicism, and Media Cynicism. The ANNALS Of The American Academy Of Political And Social Science, 546(1), 71-84. doi:10.1177/0002716296546001007
Chory, R. (2013). Differences in Television Viewers’ Involvement: Identification with and Attraction to Liked, Disliked, and Neutral Characters. Communication Research Reports, 30(4), 293-305. doi:10.1080/08824096.2013.837041
Citrin, J. (1974). Comment: The Political Relevance of Trust in Government. The American Political Science Review, 68(3), 973. doi:10.2307/1959141
Cohen, J. (2001). Defining Identification: A Theoretical Look at the Identification of Audiences With Media Characters. Mass Communication And Society, 4(3), 245-264. doi:10.1207/s15327825mcs0403_01
Chung, A., & Slater, M. (2013). Reducing Stigma and Out-Group Distinctions Through Perspective-Taking in Narratives. Journal of Communication, doi:10.1111/jcom.12050 de Graaf, A., Hoeken, H., Sanders, J., & Beentjes, J. (2011). Identification as a Mechanism of
Narrative Persuasion. Communication Research, 39(6), 802-823. doi:10.1177/0093650211408594
de Vreese, C. (2004). The Effects of Strategic News on Political Cynicism, Issue Evaluations, and Policy Support: A Two-Wave Experiment. Mass Communication And Society, 7(2), 191-214. doi:10.1207/s15327825mcs0702_4
de Vreese, C. (2005). The Spiral of Cynicism Reconsidered. European Journal Of Communication,20(3), 283-301. doi:10.1177/0267323105055259
Devasagayam, R. (2014). Media Bingeing: A Qualitative Study of Psychological Influences. In Marketing Management Association Spring Conference (pp.40-44). Chicago, IL: Marketing Management Association.
Elenbaas, M., & de Vreese, C. (2008). The Effects of Strategic News on Political Cynicism and Vote Choice Among Young Voters. Journal Of Communication, 58(3), 550-567. doi:10.1111/j.1460-2466.2008.00399.x
Feldman, S., & Sigelman, L. (1985). The Political Impact of Prime-Time Television: “The Day After”. J Of Pol, 47(02), 556. doi:10.2307/2130896
Frank, B. (2013). Viewers shouldn't think TV fiction about Washington dealings is real - The Portland Press Herald / Maine Sunday Telegram. The Portland Press Herald / Maine Sunday Telegram. Retrieved 28 July 2015, from
http://www.pressherald.com/2013/08/18/viewers-shouldnt-think-tv-fiction-about-washington-dealings-is-real_2013-08-18/
Giles, D. (2002). Parasocial Interaction: A Review of the Literature and a Model for Future Research. Media Psychology, 4(3), 279-305. doi:10.1207/s1532785xmep0403_04 Green, M. (2004). Transportation Into Narrative Worlds: The Role of Prior Knowledge and
Perceived Realism. Discourse Processes, 38(2), 247-266. doi:10.1207/s15326950dp3802_5
Green, M., & Brock, T. (2000). The role of transportation in the persuasiveness of public narratives.Journal Of Personality And Social Psychology, 79(5), 701-721.
doi:10.1037/0022-3514.79.5.701
Green, M., Kass, S., Carrey, J., Herzig, B., Feeney, R., & Sabini, J. (2008). Transportation Across Media: Repeated Exposure to Print and Film. Media Psychology, 11(4), 512-539. doi:10.1080/15213260802492000
Hart, R., & Hartelius, E. (2007). The Political Sins of Jon Stewart. Critical Studies In Media Communication, 24(3), 263-272. doi:10.1080/07393180701520991
Hawkins, S., Hoch, S., & Meyers-Levy, J. (2001). Low-Involvement Learning: Repetition and Coherence in Familiarity and Belief. Journal Of Consumer Psychology, 11(1), 1-11. doi:10.1207/s15327663jcp1101_1
Hetherington, M. (1998). The Political Relevance of Political Trust. The American Political Science Review, 92(4), 791. doi:10.2307/2586304
Hoffner, C., & Buchanan, M. (2005). Young Adults' Wishful Identification With Television Characters: The Role of Perceived Similarity and Character Attributes. Media
Psychology, 7(4), 325-351. doi:10.1207/s1532785xmep0704_2
Holbert, R. (2003). The West Wing as Endorsement of the U.S. Presidency: Expanding the Bounds of Priming in Political Communication. Journal Of Communication, 53(3), 427-443. doi:10.1093/joc/53.3.427
Holbert, R., Garrett, R., & Gleason, L. (2010). A New Era of Minimal Effects? A Response to Bennett and Iyengar. Journal Of Communication, 60(1), 15-34. doi:10.1111/j.1460-2466.2009.01470.x
Holbert, R., Tschida, D., Dixon, M., Cherry, K., Steuber, K., & Airne, D. (2005). The West Wing and Depictions of the American Presidency: Expanding the Domains of Framing in Political Communication. Communication Quarterly, 53(4), 505-522.
doi:10.1080/01463370500102228
Holbrook, R., & Hill, T. (2005). Agenda-Setting and Priming in Prime Time Television: Crime Dramas as Political Cues. Political Communication, 22(3), 277-295.
doi:10.1080/10584600591006519
Willimon B. House of Cards - Chapter One. (2013). USA.
Igartua, J. (2010). Identification with characters and narrative persuasion through fictional feature films. Communications, 35(4). doi:10.1515/comm.2010.019
Igartua, J., & Barrios, I. (2012). Changing Real-World Beliefs With Controversial Movies: Processes and Mechanisms of Narrative Persuasion. Journal Of Communication, 62(3), 514-531. doi:10.1111/j.1460-2466.2012.01640.x
Kaid, L., Towers, W., & Myers, S. (1981). Political Cynicism: a Study of "Washington: Behing Closed Doors". Social Science Quaterly, 62 (N°1).
Kleinnijenhuis, J. (2006). Negative News and the Sleeper Effect of Distrust. The Harvard International Journal Of Press/Politics, 11(2), 86-104. doi:10.1177/1081180x06286417 Lotz, A. (2007). The television will be revolutionized. New York: New York University
Press.
Weiner, M. Mad Men - Smoke Gets in Your Eyes. (2007). USA.
Masket, S. (2015). ‘House of Cards’ is the worst show about American politics. Ever… Washington Post. Retrieved 28 July 2015, from
http://www.washingtonpost.com/blogs/monkey-cage/wp/2015/03/08/house-of-cards-is-the-worst-show-about-american-politics-ever/
Matrix, S. (2014). The Netflix Effect: Teens, Binge Watching, and On-Demand Digital Media Trends.Jeunesse, 6(1), 119-138. doi:10.1353/jeu.2014.0002
McCullough, J., & Ostrom, T. (1974). Repetition of highly similar messages and attitude change. Journal Of Applied Psychology, 59(3), 395-397. doi:10.1037/h0036658 Mishler, W., & Rose, R. (2001). What Are the Origins of Political Trust? Testing
Institutional and Cultural Theories in Post-communist Societies. Comparative Political Studies, 34(1), 30-62. doi:10.1177/0010414001034001002
Moreland, R., & Zajonc, R. (1982). Exposure effects in person perception: Familiarity, similarity, and attraction. Journal Of Experimental Social Psychology, 18(5), 395-415. doi:10.1016/0022-1031(82)90062-2
Morgan, M., & Shanahan, J. (2010). The State of Cultivation. Journal Of Broadcasting & Electronic Media, 54(2), 337-355. doi:10.1080/08838151003735018
Moy, P., & Pfau, M. (2000). With malice toward all? Westport, Conn.: Praeger. Mulligan, K., & Habel, P. (2012). The Implications of Fictional Media for Political
Murphy, S., Frank, L., Moran, M., & Patnoe-Woodley, P. (2011). Involved, Transported, or Emotional? Exploring the Determinants of Change in Knowledge, Attitudes, and Behavior in Entertainment-Education. Journal Of Communication, 61(3), 407-431. doi:10.1111/j.1460-2466.2011.01554.x
Mutz, D., & Nir, L. (2010). Not Necessarily the News: Does Fictional Television Influence Real-World Policy Preferences?. Mass Communication And Society, 13(2), 196-217. doi:10.1080/15205430902813856
Myplace. (2015). Europe wide thematic analysis of young people’s views, attitudes and behaviour towards a range of social and political issues.
Nielsen.com. (2015). “Binging” is the New Viewing for Over-The-Top Streamers. Retrieved 28 July 2015, from http://www.nielsen.com/us/en/insights/news/2013/binging-is-the-new-viewing-for-over-the-top-streamers.html
Parry‐Giles, T., & Parry‐Giles, S. (2002). The west wing's prime‐time presidentiality: Mimesis and catharsis in a postmodern romance. Quarterly Journal Of Speech, 88(2), 209-227. doi:10.1080/00335630209384371
Petty, R., & Cacioppo, J. (1986). The Elaboration Likelihood Model of Persuasion. Advances In Experimental Social Psychology, 19, pp.124-162.
Romer, D. (2003). Television News and the Cultivation of Fear of Crime. Journal Of Communication,53(1), 88-104. doi:10.1093/joc/53.1.88
Schiappa, E., Gregg, P., & Hewes, D. (2006). Can One TV Show Make a Difference? Will & Grace and the Parasocial Contact Hypothesis. Journal Of Homosexuality, 51(4), 15-37. doi:10.1300/j082v51n04_02
Sigelman, L., & Sigelman, C. (1974). The politics of popular culture: Campaign cynicism and "The Candidate". Sociology Of Social Research, 58, pp.272-277.
Slater, M., & Rouner, D. (2002). Entertainment?Education and Elaboration Likelihood: Understanding the Processing of Narrative Persuasion. Commun Theory, 12(2), 173-191. doi:10.1111/j.1468-2885.2002.tb00265.x
Slater, M., Rouner, D., & Long, M. (2006). Television Dramas and Support for Controversial Public Policies: Effects and Mechanisms. J Communication, 56(2), 235-252.
doi:10.1111/j.1460-2466.2006.00017.x
Sneed, T. (2013). Is House of Cards Too Cynical About Politics? US News & World Report. Retrieved 25 July 2015, from http://www.usnews.com/news/articles/2013/01/29/is-house-of-cards-too-cynical-about-politics
Sparks, G., Nelson, C., & Campbell, R. (1997). The relationship between exposure to televised messages about paranormal phenomena and paranormal beliefs. Journal Of Broadcasting & Electronic Media, 41(3), 345-359. doi:10.1080/08838159709364412 Standard Eurobarometer. (2014). Standard Eurobarometer 82, Public Opinion in the
European Union. TNS opinion & social.
Sternbergh, A. (2014). The Post-Hope Politics of ‘House of Cards’. Nytimes.com. Retrieved 28 July 2015, from http://www.nytimes.com/2014/02/02/magazine/the-post-hope-politics-of-house-of-cards.html
Sternbergh, A. (2014). The Post-Hope Politics of ‘House of Cards’. Nytimes.com. Retrieved 28 July 2015, from http://www.nytimes.com/2014/02/02/magazine/the-post-hope-politics-of-house-of-cards.html
Tal-Or, N., & Cohen, J. (2010). Understanding audience involvement: Conceptualizing and manipulating identification and transportation. Poetics, 38(4), 402-418.
Tian, Q., & Hoffner, C. (2010). Parasocial Interaction With Liked, Neutral, and Disliked Characters on a Popular TV Series. Mass Communication And Society, 13(3), 250-269. doi:10.1080/15205430903296051
van Laer, T., de Ruyter, K., Visconti, L., & Wetzels, M. (2014). The Extended
Transportation-Imagery Model: A Meta-Analysis of the Antecedents and Consequences of Consumers' Narrative Transportation. J Consum Res, 40(5), 797-817.
doi:10.1086/673383
Wojcieszak, M. E. and Mutz, D. , 2014-05-21 "Effects of Fictional Exemplars on Political Attitudes" Paper presented at the annual meeting of the International Communication Association 64th Annual Conference, Seattle Sheraton Hotel, Seattle, Washington Online 2014-12-10.
Zajonc, R. (1968). Attitudinal Effects of Mere Exposure. Journal Of Personality And Social Psychology, 9(2, Pt.2), 1-27. doi:10.1037/h0025848
Appendixes Table1.
Items Wordings, means (M) and standard deviations for each dependent variables
ITEMS M (SD)
Trust in institutions
“There are many different institutions in your country. On these 7-points scales, please indicate how much is your personal trust in each of the following institutions:”
1. Political Parties 2.49 (1.20)
2. Government 3.04 (1.50)
3. Parliament 3.39 (1.57)
4. The head of the executive 3.08 (1.62)
Cynicism toward politicians
“On these 7-points scales, please indicate for each statement if you agree or disagree:”
1. Politicians consciously promise more than they can deliver. 5.94 (1.18)
2. Politicians are primarily self-interested. 5.12 (1.31)
3. To access positions of responsibilities, relations are more important than
abilities. 5.24 (1.43)
4. Politicians are only interested in my vote, not in my opinion. 5.31 (1.17)
Transportation
1. I could picture myself in the scenes and the event pictured by the show. 3.30 (1.72) 2. I was mentally involved in the show while watching it. 5.15 (1.41) 3. After finishing the episode, I found it easy to put it out of my mind.
(Reverse coding)
3.88 (1.54)
4. I wanted to learn how the show ended. 4.95 (1.74)
5. The show affected me emotionally. 4.19 (1.59)
6. I found myself thinking of ways the episode could have turned out differently.
3.39 (1.67)
Identification
“On these 7-points scales, please indicate for each statement if you agree or disagree:”
1. While viewing the episode, I felt as if I was part of the action. 3.33 (1.64) 2. While viewing the episode, I forgot myself and was fully absorbed. 4.29 (1.50) 3. I was able to understand the events in the episode in a manner similar to
that in, which Frank Underwood, understood them. 4.14 (1.68) 4. I think I have a good understanding of Frank Underwood. 4.45 (1.54) 5. I tend to understand the reasons why Frank Underwood does what he or
she does.
4.21 (1.75)
6. While viewing the show I could feel the emotions Frank Underwood
portrayed. 3.99 (1.69)
7. During viewing, I felt I could really get inside Frank Underwood’s head. 3.50 (1.59) 8. At key moments in the show, I felt I knew exactly what Frank Underwood
Figure1.
Model of moderated mediation
The mediated relation between exposure and trust or cynicism through transportation and identification is moderated by the extent of prior exposure and the intensity of prior exposure.
H1
Extent / Intensity of Prior Exposure Trust / Cynicism Transpor-tation Identifi-cation Narrative H2 a H2b H3a-‐b H4a-‐b H5a-‐b
