You lying press make me so angry I got to do something! How hostile media perceptions drive political participation.
Laura Schacht 10602097 Master Thesis
Graduate School of Communication Communication Science (Research Master)
University of Amsterdam Supervisor: Dr. Linda Bos
Abstract
This quasi experiment, investigating the relation between hostile media perceptions and political participation is joining a growing research body that demonstrates consequences of biased media perception. The research is embedded in an issue of increasingly negative attitude towards immigration and refugees in Germany and the rise of populist parties. Besides the direct effect the underlying mechanisms are of interest in this study and conditional effects as political efficacy and attitudes are tested, as well as mediation via emotions; namely anger and disgust. The results support previous research and indicate a main effect as well as an indirect effect; anger promotes participation. Furthermore this study suggests that disgust is not treated as emotion, but felt body state. No conditional effects are found to be significant.
Keywords: Media effects, hostile media perception, political participation, emotions,
You lying press make me so angry I got to do something! How hostile media perceptions drive political participation.
“Lügenpresse” (the ‘lying press’) – an expression used as buzz word during the Second World War by representatives of Nazi Germany to defame critical news reporting – underwent a boom in Germany recently. The ‘patriotic Europeans against the Islamization of the Occident (Pegida) dusted it off and brought it back as an often, nearly inflationary used word in Germany, in fall 2014. Consequently it won ‘non-word’ of the year (SZ.de/cag, 2015). Apart from its shady origin, it reflected the apparently smoldering discontent that obviously many people felt towards media in Germany. In the case of Pegida, people seemingly perceived a threat by the
‘Islamization’ through Muslim immigrants and refugees. Some of the people turned into enraged citizens (German: “Wutbürger”), who – particularly in Dresden – went on protest marches to announce that they felt betrayed by the German government, the so-called “Volksverräter” (traitors of the people). Hand in hand with the rise of this movement, violent attacks towards immigrants and refugees increased drastically. According to press reports attacks increased from 33 assaults in the three months before the first Pegida march, to 76 in the three months after the first march (Deutsche Presse Agentur, 2015).
Wolfgang Donsbach, Professor of communication sciences in Dresden (2014) names one possible reason of the accusations towards the media: disenchantment with media: a loss of both, credibility of, and trust in, established media reports. Although many are shouting ‘Lügenpresse’, there are also people that have not called media a ‘lying press’, although content of the news reporting has not changed, but remained exactly the same. This bias in perceiving the position of the news in relation to one’s own position is called ‘Hostile Media Perception’ (HMP) in political and communication science. In 1985, Vallone, Ross and Lepper (1985) demonstrated this biased
perception for the first time among groups of pro-Israeli and pro-Arab partisans on the reporting of the Beirut massacre. In both groups, the coverage of the massacre, as well as the people responsible for the reports, were evaluated as biased against their side by participants. So while a neutral audience would perceive a neutral article as balanced, members of the audience that are not neutral, but biased towards one of the sides, will not see the article as neutral. The
phenomenon itself has been replicated thoroughly in academic research (Feldman:2014hh and Perloff, 2015).
Much more relevant – in respect to recent developments in Germany that go hand in hand with a rising support for populist parties and party leaders – are the consequences of a perceived bias in the media. To date only a few studies have been looking at behavioral and attitudinal effects of HMP (Feldman, Hart, Leiserowitz, Maibach, & Roser-Renouf, 2015; Gunther,
Christen, Liebhart, & Chia, 2001; Tsfati & Cohen, 2005) but results suggest that perceived media bias can have extensive influences on opinion, attitudes and behavior in the political arena. Taking the relation between HMP and political participation as starting point for this study, questions arise about the mechanisms that underlie this relationship. A better understanding of those mechanisms would open up ways to reduce the biased perception (Perloff, 2015). Firstly research so far has been missing out on individual differences that shape the effects of HMP on behavior(Feldman, 2014), secondly perceptual theories and concepts would benefit from understanding the mechanisms underlying the effects to improve concepts about conditional effect on, for example, the spiral of silence (Scheufele & Moy, 2000). Hence this study is not only testing the impact of HMP on participation, but also trying to look for individual differences, and the underlying mechanisms. All this is embedded in the German context, and by reference to immigration and refugees.
A major aim of this research, therefore, is to improve validity of previous research on the consequences of HMP on political participation, including political efficacy and anger, as
conditional and mediating variables. As well as this, new variables are introduced into the model to investigate the mechanisms underlying the effects more in-depth. This leads to the following research question: To what extent, and under which circumstances, does a perceived media bias affect participation in the political arena?
Theory
HMP, synonymously used in literature to hostile media effects is the response of a person to consumed media content. This view on HMP also entails that the audience is active, rather than passive, and responses vary differently to media content, and therefore, the individual interpretation of media content gains more importance than the message itself (Gunther & Schmitt, 2004). This does, of course, not rule out the possibility, that this biased perception can have effects on, for example, perceived public opinion, political communication and behavior (Perloff, 2015); but most previous literature actually analyzed how HMP was triggered, not what consequences it entailed. Literature on consequences is still rather scarce, leaving many
questions unanswered about conditional effects, indirect routes, and effect strength.
Throughout the 30 years since the HMP has been scientifically examined for the first time, further research using experimental and survey methods has replicated Vallone et al.’s (1985) results and found support for its existence throughout many different issues, such as primates in laboratory research (Gunther et al., 2001), genetically modified food (Gunther & Schmitt, 2004), medical practices (Schulz, 2013) and the Sarajevo market bombing in 1994 (Matheson & Dursun, 2001). A meta study on the HMP by Hansen and Kim (2011) that put 34 studies under analysis, observed HMP to have a moderate, yet robust effect on perception across
media formats, study designs, and even for respondents who were less involved in the topic – even though they were less affected than those who were actually more involved. Nevertheless effects were significant, and found across several studies.
In the classic definition of the HMP, the biased perception was evenly distributed on both sides and the neutral participants did not indicate any bias in the news reporting. Since it is rather impossible for journalists to always write news reports that are either totally neutral, or perfectly balanced regarding the arguments, Gunther et al. introduced the term “relative hostile media perception” (2004, p. 57). The relative perception takes into account, that even for genuinely slanted news, the HMP still occurres on both sides in relation to the ‘center’. In deference of this slant, the relative HMP was taken as basis for this study. Hence, even for a slightly leftish news report, there would have still been participants that saw the article as neutral and / or balanced, and therefore would place it on the same ideological position as themselves. In the case of the relative HMP, these people do not perceive any HMP, while there are still other consumers of the article that will perceive a bias, no matter if they perceive this bias to be to the left or to the right, it will differ from their own position.
Implications for Political Behavior
As mentioned in the introduction, an emerging body of literature supports the notion that HMP has considerable influences on the political arena (Feldman et al., 2015; Ho et al., 2011; Rojas, 2010). Democracies build on active citizens (Fishkin, 1997), and for those with a perceived media bias, it has been found to have an effect on participation.
Since HMP is a response to information sources, it seems to only occur when the message(s) that provoked it had a high reach, while there was no perceived bias found in messages that were said to have a low reach (Gunther & Liebhart, 2006). The necessary
condition of a perceived high reach of the news source was suggested to lead to a perceived difference between one’s own opinion, and public opinion. In the theory of the spiral of silence by Noelle-Neumann, this incongruence affects the willingness to express one’s own opinions in public negatively (Scheufele & Moy, 2000). Perloff (1989) however argued that this might only count for lowly involved people. On the other hand, highly involved individuals that felt the need of justice and that see manipulation by political elites, might not turn silent, but speak out loud, and eventually participate in the political arena (Hwang, Pan, & Sun, 2008). These results are backed by Matthes, Morris and Schemer (2010), where support for the spiral of silence was only found in individuals that had low or moderate attitude certainty.
Previous research showed mixed results for the main effect of HMP on political
participation, but it has to be mentioned that participation was measured very differently in each of the studies. Feldman et al. (2015) measured climate change activism with two items, asking for the frequency with which they had engaged in two actions over the last 12 months, and found a conditional effect by ideology. While no direct effect was found in Ho et al.’s (2011) general participation index, where only an indirect (and negative) effect via trust and political efficacy could be demonstrated, they revealed a significant positive main effect of HMP on issue specific participation, with data derived from a different data set. Significant positive main effects were found by Rojas (2010) on offline and online ‘corrective’ actions. Besides obvious variance in the results regarding the main effect of HMP on political participation, it became evident that
measures for the dependent variable differed across studies, and it therefore seemed very important to thoroughly rethink the measures to ascertain political participation.
Resuming previous research on the relation between HMP and political participation, it became evident that there must be a more complex mechanism behind this relation and that HMP
alone does not / or not alone predict political participation. Nevertheless, and with the aim to shed more light into a field where research has found evidence for contradictory hypotheses, a positive main effect from HMP on political participation – like Rojas (2010) found support for – was expected (H1).
In accordance with Valkenburg and Peter (2013), this study addresses one of five challenges that media-effects research faces today: the call for more conditional effects; not all effects hold for all people, but there are certain variables that have the power to explain why effects are stronger for some people but not others. And “only by formulating clear hypothesis about which individuals are particularly susceptible to the effects of media are we able to specify the boundary conditions for media effects” (Valkenburg & Peter, 2013, p. 203).
A variable that found entrance in many studies explaining political participation – besides classical SES variables (Ansolabehere, Iyengar, Simon, & Valentino, 1994; Brady, Verba, & Schlozman, 1995; Feldman et al., 2015; Theiss-Morse & Hibbing, 2005; Valentino, Gregorowicz, & Groenendyk, 2008) – and that could take the place of this conditional variable, was political efficacy. Internal political efficacy is a psychological construct that describes the perceived competence in the political arena; of the participant itself, and the responsiveness of the government (external political efficacy) (Beierlein, Kemper, Kovaleva, & Rammstedt, 2012; Campbell, Gurin, & Miller, 1954). For the purpose of this study, it was chosen to look at internal political efficacy only, a concept that refers to the ability to understand and actively participate in politics, and was found to be related to political participation among others (Craig, Niemi, & Silver, 1990). Although both types of efficacy have been found to affect political participation (Niemi, Craig, & Mattei, 1991; Verba & Nie, 1987), internal efficacy represents a personality trait and therefore is more stable, while external efficacy is not (Beierlein et al., 2012).
Drawing on the same theoretical arguments as for the direct effect; that involved people will try to rally their opinion and convince others after fearing that public opinion is against their position, people should be more likely to actually speak out if they feel personally competent enough to do so in the political arena (Matthes et al., 2010). Rendering the results and
recommendations of previous research, the following hypothesis was developed:
H2: Internal political efficacy positively moderates the impact of HMP on political participation.
Emotions at work. When we think of emotions typical examples that immediately cross our minds are very likely fear, joy, love or anger. In social psychology, these basic emotions have been used to explain social behavior for decades, but recently, political science started to
understand the important effects of emotions in the political arena (Ryan, 2012). Newest research in neuroscience has found support for emotional responses to political content as changes in cortisol and testosterone levels, or the activation of emotion related brain regions as the
amygdala, that have been found in relation with voting (Jiménez, Aguilar, & Alvero-Cruz, 2012; Kivlighan, Granger, & Booth, 2005; Mehta & Josephs, 2006; Neiman et al., 2013; Rule et al., 2010), but also partisan identification and evaluation of politicians (Blanton, Strauts, & Perez, 2012; Knutson, Wood, Spampinato, & Grafman, 2006). Such core evidence cannot be rejected: obviously our emotions play an important role in political decision mmaking and behavior, led, for example, by hormones, and activated brain regions, that we cannot necessarily influence and control directly.
The theoretical mechanism that is expected to underlie the relation between HMP and emotions – and eventually political participation – is the cognitive appraisal theory (Lazarus, 1991). According to this theory, emotions are dependent on an appraisal of a certain situation.
Depending on, for example, how unfair a situation is for us, on the source of this unfairness, and how well we are in control of this situation, our appraisal of the situation will differ – and with it our emotions (Lazarus, 1991; Yap & Tong, 2009).
Two emotions that have been put in the focus of political research have been anger and fear, although only anger and other related emotions as indignation led to significant changes in behavior, while fear resulted in conflicting evidence. (Brader, Valentino, & Suhay, 2008; Hwang et al., 2008; Valentino et al., 2008; Valentino, Brader, Groenendyk, Gregorowicz, & Hutchings, 2011). As Newhagen (1998) points out in his experiment testing approach-avoidance and memory, anger “is the most involving and invigorating of the negative emotions” (Newhagen, 1998, p. 266) that can result in an attack or approach towards the source of anger. While anger activates a more approaching behavior, fear activates a flight mode. The third emotion under observation in Newhagen’s experiment is disgust; a response to a less threatening, but still evident danger (Newhagen, 1998), and has before been called the emotion of avoidance. The results showed a significant difference between the approach and avoidance ratings of images that induced anger, fear and disgust; where anger scored highest on approach ratings, while disgust scored highest on avoidance ratings (Newhagen, 1998). In order to make the model not too complex and since fear scored moderate between anger and disgust (Newhagen, 1998), only anger and disgust are included in this study.
In the cognitive appraisal theory both emotions belong to the category that is provoked by harms, losses and threats (Lazarus, 1991). Those negative emotions should therefore occur for participants that somehow evaluate the media reporting as somewhat threatening, a description that suits well on hostile media perception. A perceived public opinion that is contradictory to the
participant’s opinion – reflecting hostile media perception – should leave this person threatened, and should therefore result in negative emotions.
In light of these, and previous results, anger and disgust were included as the mediating variables, leading to the following hypothesis:
H3: HMP leads to more anger, which in turn, leads to an increase in intentions to participate.
Disgust in turn is hypothesized to have a negative effect on participation, an expectation that builds upon the results of Newhagen (1998), but hasn’t been tested so far on the effect on political participation:
H4: HMP leads to more disgust, which in turn leads to less participation.
Looking at hypothesis three and four and their theoretical reasoning, it becomes clear that there must be something remaining to explain since HMP is expected to increase both emotions, but the emotions are theoretically expected to have contradictory effects on participation. Again, referring to Valkenburg and Peter (2013) there might be conditional effects that play an important role; in this case defining under which conditions a participant feels (more) anger or (more) disgust as emotion.
The theory of principal of consistency predicts that we would expect behavior that is consistent with attitudes; nevertheless already in 1934, La Pierre concluded that attitudes do not reliably predict behavior (LaPiere, 1934). Attitudes have been defined as “a psychological
tendency that is expressed by evaluating a particular entity with some degree of favor or disfavor” (Eagly & Chaiken, 1993, p. 1) and in the background of this study favor or disfavor towards immigration and refugees.
Equally to the HMP, attitudes are subjective and evaluative, and can be more or less conscious, and peoples’ ideologies do not exist in isolation from attitudes (Maio, Olson, Bernard, & Luke, 2006).
To include attitudes into the model has never done before, but previous research in neuroscience revealed evidence for an interplay of ideology, attitudes and emotions (Kaplan, Freedman, & Iacoboni, 2007). In an FMRI experiment, Kaplan et al. (2007) investigated how political party affiliation and political attitudes modulate neural activity. Emotional sub regions of certain parts of the brain were shown to be activated by viewing opposing political candidates and where found to be modulated by political attitudes (Kaplan et al., 2007). Assuming that the viewing of party members of opposing parties results in a similar state to the consumption of opposing political news, and looking at the brain regions activated, the findings could be transferred to this experiment.
Although no previous research has been conducted investigating attitudes moderating the effect of HMP on emotions, I expect that strength of attitudes towards immigration and refugees facilitates the positive effect of HMP on anger; that people that perceive a biased news coverage conceive more anger if they have strong attitudes towards the issue, but less anger when their attitudes are low or moderate. People with stronger attitudes towards the issue are equally expected to conceive more disgust, and consequently participate even less. Since relations have not been explored yet less predictive research questions are formulated that leave room for exploration of the data:
RQ2: Will there be a positive indirect effect of HMP on participation via anger that will be conditional on the strength of attitudes towards immigration and refugees?
RQ3: Will there be a negative indirect effect of HMP on participation via disgust that will be conditional on the strength of attitudes towards immigration and refugees?
Methods Design and Procedure
The respondents for this quasi experiment were approached via Facebook and other online discussion groups and forums, in a convenient snowball sampling. A link underneath the cover story and invitation redirected them to the online questionnaire, administered using Qualtrics. Respondents were asked to read a newspaper article, and subsequently answer certain questions regarding ‘media and politics’.
From the total amount of 227 people that clicked on the link, 51.5% finished the
questionnaire. Two of them had to be deleted from analysis due to their answer to the seriousness check that indicated that they only clicked through the questionnaire without answering or reading the questions consciously and honestly (see appendix A1), 16 were not included in the analysis due to missing values on the dependent variables. The remaining 99 participants were used for the analysis. From the total of 99 cases that were left for the analysis, 42.4% were female (3% went for the ‘no answer’ option), a number that did not lie very close to the German population with 50.9% women though (Statistisches Bundesamt, 2014) and the average age in the sample was 35.5 years (SD = 13.93) with the youngest participant with 20 years, the oldest with 80. The mode for education was at higher education entrance qualification, the category that was most frequently chosen with 36 times, followed by masters degree with 28 times. The level of education was therefore much higher for the sample, compared to the population; while higher education entrance qualifications was already higher with 36.4%, compared to 27.9% in the population, 28.3% have a master degree in the sample, while only 8.3% in the German
population have a university degree (Statistisches Bundesamt, 2013). Hence, lower educational levels were underrepresented.
The cover story was a “yearly report on media and politics in Europe” (see appendix A1) and the newspaper article was reported as being a common article found like this, or very similar to this, in all of the German newspapers during the last six months. The article (see appendix A2), titled “Stress test of the Nations” had 364 words in total and included equal amounts of positive and negative arguments towards immigration and refugees. It was put together from eight newspaper articles that were also used for the quantitative content analysis of media coverage and frames on the human tragedy between 2012 and 2015 on Lampedusa from De Swert, Schacht and Massini (n.d.). To pretest the balance of positive and negative arguments the German coders for the same content analysis – who have read each more than 120 articles on that issue – were asked to mark positive and negative connoted sentences, and identified each six sentences as negative and had a 83.3% overlap for positive sentences in the final version of the article.
Hostile media perception.
The independent variable in this study is the perceived bias in the news report. Previous measures were adapted and resulted in two items: general HMP, and article specific HMP. The general HMP consisted of the difference between the participants’ position in the political spectrum, and their perceived position of the media in general (Rojas, 2010). Both questions could be answered on a scale from “very liberal / left” (-5) to “very conservative / right” (+5) and the measure of difference, the final scale for HMP resulted in a mean of 2.64 (SD = 2.39),
indicating that the position of the participants themselves (M = -1.39, SD = 2.06) differed from the perceived media position (M = .21, SD = 2.08); this difference was significant (t(79) = -3.83.
p < .001). The means indicate that the participants were more leftist, and media was seen slightly
rightist on average.
The second HMP measure was also a difference measure, like the previous item, but contrary to the previous one, it was article specific. Participants were asked to indicate to what percentage they think the article is negative, neutral and positive (Gunther & Schmitt, 2004). The significant differences (t(98) = 3.09. p < .01) between negative (M = 31.28; SD = 22.12) and positive (M = 20.86, SD = 18.54) percentage was calculated and used as final measurement (M 25.92, SD = 23.74). Although both measures did not indicate any direction of biased perception, respectively if the person placed itself more on the right or on the left side for ideology, the differences allowed assumptions about the intensity of the HMP, respectively how much the perceived media ideology differs from one’s own.
Dependent Measures
Political participation. The main dependent variable was intended behavior, and the scale was created by merging and modifying the issue specific participation scale from Ho et al. (2011) and the activism and radicalism intention scales from Moskalenko & McCauley (2009). Participants were asked on a seven point Likert scale (ranging from 1 = not likely at all, to 7 = very likely) how likely it is that they would participate in one of the following actions regarding immigration and refugees. Possible answer categories reached from joining a public
demonstration, to supporting an organization that might break laws while fighting for one’s own sentiment. (See questionnaire in appendix A1 for more answer categories).
To extend the intended behavior scale, two additional measures were implicated to evaluate the willingness to participate. Participants were asked to select items of interest, as many as wanted, from seven presented options; the participants could choose from seven
petitions to sign after the questionnaire, and seven Facebook groups to follow. The options where presented in random order and showed either neutral options (neutral towards the issue of
immigration and refugees) or equal amounts of positive / negative options towards the issue (see appendix A1).
Due to different scales all items were standardized. Initially all 10 items were used for a CFA, the Kaiser-Meyer-Olkin measure of sampling adequacy was .81 and therefore above the recommended threshold of .6 and significant test for sphericity (2
(45) = 526.84, p < .001). The initial Eigenvalues showed that the first factor explained 45.89% of the variance, the second 12.3%, and the third 10.87%. All items remained; their contribution to the structure was given. The first factor clearly contains all intentions, but also both direct participation items are close to a .5 value. A reliability analysis supports the finding, and the final scale consists of all 10 items ( = .87) with positive values indicating an intention to participate that is above the average (M = .00, SD = .67).
Political Efficacy. The scale to measure political efficacy was previously tested in the German language context by Beierlein et al. (2012) and reached sufficient level of reliability. Internal efficacy was operationalized in two questions, each on a seven point Likert scale. The choice of seven point over five point scales is conclusively justified by the preference of
consistent measurement scales throughout the questionnaire. reached .78 for internal efficacy with a mean of 5.43 (SD = 1.07), which is relatively high, above the neutral category (Field, 2013); so on average, people agreed somewhat that they are good at understanding and assessing important political issues and that they have the confidence to take active part in a discursion about political issues.
Emotions. The respondents were asked to indicate if and to what extent they experience the emotions during and / or after reading the newspaper article. The mediating variables anger and disgust were measured on a seven-point answer scale (1 = fully agree, to 7 = fully disagree). With a mean of 4.33 (SD = 1.65) participants felt more anger than disgust (M = 2.52, SD = 1.82).
Strength of attitude towards immigrants and refugees. Attitudes towards the issue were measured with 6 statements that respondents were asked to agree / disagree with on a seven-point Likert scale. The items were created from main stereotypes found in the content analysis from De Swert et al. (n.d.) and were either pro immigration and refugees: “Someone who risks their life on the way here should be welcomed with open arms”, or negatively assigned: “Border controls should be enforced to stop refugees before they enter Europe”. Negative statements were reversed, and all items were recoded into strength of attitude, whereas the middle value (‘I am somewhat in the middle’) was coded as lowest value, the extremes (‘I fully agree’ / ’I fully disagree’) were the highest value. A confirmatory factor analysis with all 6 items was conducted, Kaiser-Meyer-Olkin measure of sampling adequacy was .76 and the test for sphericity significant (2
(45) = 526.84, p < .001). The first factor explained 43.48% of the variance; the eigenvalue of the second factor explained 17.46% of variance. All items loaded higher than .5 on the first factor, the reliability analysis supported the finding and the final scale consisted of all 6 items ( = .72, M = 2.48, SD = .7).
Controls. Standard control measures were age, income (Mode = 1201-2000€ a month), gender and education, next to ideology (M = -1.39, SD = 2.06), frequency of media use and institutional involvement, where 39 participants indicated to work voluntarily in an organization. In previous studies the latter was identified together with time, income and self-described skills, to predict political participation better than the standard SES model (Brady et al., 1995).
All hypothesis were tested with the PROCESS macro for SPSS developed by Hayes (2012), which, similarly to SEM, can estimate complex models including moderation and mediation, but requires less cases. All models used 10.000 bootstrap samples and bias corrected and accelerated 95% confidence intervals (bca).
Results Test of Hostile Media Perception
To prove the general assumption of the Hostile Media Perception: that people advocating a more extreme ideology, perceive media as more hostile, a simple independent t-test revealed the expected differences: the group of participants that scored around the center of the political ideology scale between -1 and 1 (M = 1.28, SE = 1.13) perceived media as less hostile than participants that scored more towards the extremes of the political spectrum between /+ 2 and -/+ 5 (M = 3.09, SE = 2.53). This difference was significant t(97) = -3.45, p < .01. Nevertheless, this only held for the general HMP, which was a scale that runs from 1 to 7. The article specific HMP, a scale that ran from 0 to 100 showed no significant differences (t(97) = -.66, p = n.s.) between people around the center of the political spectrum (M = 23.2, SD = 18.31) and the people with moderate to high values (M = 26.85, SD = 25.36). Subsequently only the general HMP scale was used as independent variable for following analysis.
Hypothesis
To test the first hypothesis of a main effect from HMP on political participation a linear regression model was estimated, including independent and dependent variable and all controls, namely: political ideology, frequency of media use, income, age, gender, education and
organizational involvement. The results indicated that HMP and the controls explained 26.4% of the variance (R2 = .264, F(8, 87) = 3.903, p = .001), the direct effect of HMP on participation
was significant (b* = .06, t = 2.25, p < 05, 95% CI .01, .11), also the controls ideology (b* = -.08, t = -2.83, p < 01, 95% CI -.15, -.03) and frequency of media use (b* = .12, t = 2.09, p < 05, 95% CI .01, .23) had significant effects. Income (b* = -.03, t = -.58, p = n.s., 95% CI
-.11, .06), age (b* = .00, t = -.3, p = n.s., 95% CI -.01, .01), gender (b* = .02, t = .18, p = n.s., 95% CI -.22, .26) education (b* = -.03, t = -.73, p = n.s., 95% CI -.13, .06) and organizational involvement (b* = .19, t = 1.38, p = n.s., 95% CI -.09, .47) were not significant.
Opposing to the assumptions hypothesized in the first moderation hypothesis, internal political efficacy had no conditional effect on the relation between HMP and PP ( = .062, SE = .031, p = n.s., bca 95% [.000, .124]). In this simple moderation model (model 1) by Hayes
(Hayes, 2013, p. 442) ideology and frequency of media use were included as control variables due to significant effects on the outcome variable. This model also did not show any significant main effect of HMP on PP ( = .053, SE = .036, p = n.s. bca 95% [-.019, .125]). With no main effect nor moderation hypothesis 1 was rejected and internal political efficacy was excluded from the model: there were no significant differences between participants with higher levels of
internal political efficacy and participants with low or moderate levels of internal political efficacy of HMP on the intention to participate in the political arena.
Hypothesis 3 and 4 incorporated two mediators: anger and disgust, evoked by the consumption of the article that was presented in the study. While the third hypothesis predicted that increased levels of anger for participants with higher perceived media bias would result in an increase in the intention to participate, disgust was hypothesized to decrease intended
participation. PROCESS model 4 was estimated, including two mediators: anger and disgust; and the covariates.
Table 1. OLS Regression Coefficients, Standard Errors and Model Summary Information for the Multiple Mediator Model
Consequent
M1 (ANGER) M2 (DISGUST) Y (POL PARTICIPATION)
Antecedent Coeff. SE p Coeff. SE p Coeff. SE p
X (HMP) a1 .166 .068 017 a2 .217 .075 .005 c’ .047 .024 .055 M1 (ANGER) --- --- --- --- --- --- b1 .171 .036 < .001 M2 (DISGUST) --- --- --- --- --- --- b2 -.009 .033 .793 C1 (IDEOLOGY) f1 -.105 .079 .188 f3 -.020 .087 .817 g1 .047 .024 .003 C2 (MEDIA USE) f2 .076 .135 .577 f4 .083 .149 .578 g2 .126 027 .932 Constant iM1 3.394 .681 < .001 iM2 1.525 .749 .045 iY -1.488 .256 < .001 R2 = .086 F(3, 95) = 3.9, p < .05 R2 = .087 F(3, 95) = 3.029, p = .033 R2 = .400 F(5,093) = 12.988, p < .001
The coefficients and the model summaries for the multiple mediation model were condensed in Table 1. The data showed support for the third hypothesis; HMP positively predicted anger, which in turn predicted the outcome variable political participation. Hence mediation occurred for anger; the indirect effect via anger was significant ( = .028, SE = .014, bca 95% [.004, .060]), the population value is unlikely to be zero, the bootstrapped confidence interval was positive. Participants that scored higher in their perceived media bias therefore were more likely to feel angry, and their intention to participate increased subsequently.
Nevertheless, the mediation for disgust was not significant; HMP had a significant effect on disgust as emotion, but disgust had no significant effect on participation. Besides the indirect effect of HMP on political participation via emotions – in this case anger only - the direct effect remains close to significant, albeit small.
So far the results supported the proposed mediation via the emotion anger, which amplifies the intention to participate. This mediation variable anger was assumed to be
moderated by the strength of attitude towards the issue – immigration and refugees – in RQ2. It was thought that the positive relationship between HMP and perceived anger would be stronger for participants that have a stronger attitude towards immigration and refuges than among participants that have a weak attitude. Conditional mediation model 7 (Hayes, 2013, p. 447) was estimated, predicting political participation with HMP via the two emotions anger and disgust, that were expected to vary under different conditions of the attitude towards the issue. Ideology and frequency of media use were again included as control variables. The main effect of HMP on political participation remained equal to the previous analysis (Table 1), only the coefficients for the mediators changed. Consistent with the multiple mediation model, HMP predicted anger ( = .143, SE = .070, bca 95% [.004, .282]) and disgust ( = .218, SE = .078,
bca 95% [.064, .373]) significantly. Nevertheless, there was no interaction effect of HMP by attitude strength on both emotions (Anger = -.1260 SE = .087, bca 95% [-.299, .048];
Disgust = .010, SE = .097, bca 95% [-.183, .202]).
The nature of the relationship did not significantly differ under a change of attitude strength, nevertheless minor differences were significant in the interaction term for low attitude strength ( = .040, SE = .018, bca 95% [.010, .082]) on anger, all other effects remained
insignificant.
Research question three, that suggested an increase of disgust for participants that have a strong attitude towards the issue and subsequently engage less had to be negated, due to no significant differences whatsoever. Neither was the mediation effect significant in the model estimated before (Table 1), nor was the interaction effect significant.
Overall, a main effect of HMP on intentions to participate seemed likely, but significance level failed to drop under the of .5 in every model besides the regression model. A mediating effect of anger was supported by the data. Nevertheless, neither internal political efficacy, nor the strength of attitudes towards immigration and refugees conditioned any of the effects predicted by HMP on political participation
Discussion
Taken together, the main effect of HMP on political participation is somewhat, and a mediating effect of anger is supported by the data. Hypothesis one and three found support in the data, hypothesis two and four have to be rejected: no significant support for the moderating effect of internal political efficacy on the main effect, and no evidence of mediation via disgust, is found. Furthermore, the research questions two and three have to be negated: the strength of attitude does not moderate the effect of HMP on emotions.
Although not all hypotheses can be supported by the data, this study helps to validate previous research on the mechanism underlying the effect of biased media perception on political behavior. Previous results of a direct effect can be somewhat supported, additionally this study also adds an explanation for the mechanism, namely via the emotion anger. These results go in line with previous research that included anger as key mediating variable (Hwang et al., 2008; Valentino et al., 2008) across different contexts and issues, and reflects the difficulties it bears to include conditional effects as Valkenburg and Peter (2013) recommend.
Regarding the study itself, there are some limitations to mention. First, the assignment to the independent variable manifestations was not random, as is required to make clear causal conclusions (Imai, Tingley, & Yamamoto, 2013), but depended on attributes of the participant that were not controlled for in this quasi experiment. That causal inferences are difficult to be drawn is especially the case if the intermediate variables also can not be manipulated, as in this study (Imai et al., 2013). Inferences about causal relations have therefore to be made cautiously. While the use of a quasi experiment results in a downside for causal inferences, it does improve external validity of the experiment contrary to lab experiments for example. Besides no random assignment, the used sample was gathered as convenience sample on Facebook and other political forums, which might have led to the slightly leftish ideology among participants, and generally quite high participation, since it can be expected that people that are active on political forums and participate in a survey on politics also participate more in the political arena in general (Brady et al., 1995). Another point that limits the interpretational reach of the results, is the single exposure of the participants to one news report at one time point only. Some effects might be hidden under weak manipulations, and others might vanish very quickly. This is aggravated by the fact that the exposure was forced, and does not represent the normal selection
a person would necessarily make in a normal media environment. This can lead to an overestimation of effects.
The general limitations have to be kept in mind in the following part where the single hypotheses are discussed. The second hypothesis is not supported by the data; the moderation effect of internal political efficacy on political participation is not significant. Although previous research has shown that activism was affected by high levels of political efficacy (Ansolabehere et al., 1994; Brady et al., 1995; Feldman et al., 2015; Ho et al., 2011; Valentino et al., 2008), this data does not support these findings. Since the measurement has been tested before in the
German context, it is unlikely that the measurement scale lacked validity, it is more likely though that the sample lead to a ceiling effect, which occurs when many participants score near to the maximum (Hessling, Traxel, & Schmidt, 2004). Looking back at the data this could very likely be the case for political efficacy with an average level of 5.3 (SD = 1.17). The reference values that Beierlein et al. (2012) give in their evaluation of the measures for political efficacy are never higher than 4.1 on average age and for all levels of education. With more than 1000 participants and a representative sample (Beierlein et al., 2012) I suggest that high values in this sample are likely a result of the sample itself, but effects could be found if the sample would be improved. Nevertheless this study could not accomplish this.
The mediations that were hypothesized were only found for anger, but no significant effects were found for disgust. That there are no mediation effects via disgust is contrary to expectation from previous research. In the study of Newhagen (1998) disgust lead to the highest avoidance, while anger lead to the highest levels in approaching. Two possible explanations can be used: The first lies in the statistical closeness of disgust to the concept of anger; both correlate significantly (r = .411, p < .001). But instead of having no effect, this explanation would suggest
that disgust is affected in the same way, just not measuring two independent constructs but the same. The distinction between anger as emotion and disgust as felt body-state, that Zinck and Newen (2008) make, seems to be more plausible in this case. Felt body states are classified as automatic reflex, responding to organism-threatening conditions, and need less complex
behavioral programs than emotions (Zinck & Newen, 2008). Hence disgust could be a felt body-state, an automatic reflex that is triggered by the threatening situation of immigration and refugees, and could lead to a simple flight mode, but not lead to any effects on participation subsequently.
The last test, that analyzed the interaction term of HMP and strength of attitudes on emotions showed no significant effects either. Hence strong attitudes did not enhance anger, weak attitudes did not result in less anger. To include attitudes into the model has never been done before in an online quasi experiment, and although theoretically effects are expected the data does not show any. The effects Kaplan et al. (2007) found, are first of all very complex, and it would be possible that they can not be transferred to an experiment like this so easily. So far this combination of political content, attitudes and emotions has not been tested outside of neuropsychology, and the processes might also not be conscious enough to test them in self-evaluated questionnaires. Regarding the two final research questions, if moderation occurs, the results suggest to clarify concepts and mechanisms before they are applied in an already complex relation of HMP and participation.
The results on the main effect and mediation via anger, contribute to a growing body of literature that suggests that emotions play a crucial role in the relation between HMP and political participation, and HMP as a consistent predictor of participation (Hwang et al., 2008; Rojas, 2010; Valentino et al., 2008; 2011). In line with Fishkin’s Deliberative Democracy
(Fishkin, 1997; Thompson, 2008) the effects I found should enhance a peaceful conflict resolution, since HMP activates citizens to take part in the political discourse. Even though a negative emotion mediates the effect, as long as it increases the willingness to participate it is seen as promising for a deliberative democracy. In spite of this, it also has to be mentioned that many citizens will not perceive bias in the media, while others – citizens with more extreme ideologies – will very likely perceive this bias, which in turn results in political participation but in polarizing instead of deliberative way (Rojas, 2010). Hence, Fishkin’s concept of democracy might work under ideal conditions, in the case of Germany it seemed as if it had failed: the neo-nationalistic movement gathers pace. According to Lobo (Lobo, 2015) it became socially acceptable to yell out loud “Germany first” (German translation: “Deutschland zuerst”). One participant wrote “Abschieben” and “Volksverräter” (“deport” and “traitors of the people”) in an open answer field in the questionnaire. Facebook comments and public accusations towards the immigrants and refugees reach a new high in contempt for mankind (Lobo, 2015)With the intensification of the extremes, fuelled by a perceived media bias, the extremists call for an anti-establishment and they share their populist philosophy – in this case nationalistic – with
everyone keen, or not keen, to hear it. In extreme cases it seems to be a spiral of HMP, emotions and participation in the political and public arena that pushes itself further up to the even more extremes. The rise of populist parties has not yet reached it’s high and according to Mudde (2015) it will not decay.
Results show that the consequences of HMP have to be put under further analysis to understand not only the effects but also the mechanisms better. Only when we fully understand who perceives HMP (antecedents of the variable) and for whom and under which conditions it leads to behavioral changes, politics and society can react to this spiral. In the case of this
particular research, Hostile Media Perception does lead to increased political participation, and anger plays a crucial role in this relation.
References
Ansolabehere, S., Iyengar, S., Simon, A., & Valentino, N. (1994). Does attack advertising demobilize the electorate? American Political Science Review, 88(04), 829–838. http://doi.org/http://dx.doi.org/10.2307/2082710
Beierlein, C., Kemper, C. J., Kovaleva, A., & Rammstedt, B. (2012). Ein messinstrument zur erfassung politischer kompetenz- und einflussüberzeugungen. A measurement instrument for political efficacy. Presented at the gesis, Köln.
Blanton, H., Strauts, E., & Perez, M. (2012). Partisan identification as a predictor of cortisol response to election news. Political Communication, 29(4), 447–460.
http://doi.org/10.1080/10584609.2012.736239
Brader, T., Valentino, N. A., & Suhay, E. (2008). What triggers public opposition to immigration? American Journal of Political Science, 52(4), 959–978.
http://doi.org/10.1111/j.1540-5907.2008.00353.x
Brady, H. E., Verba, S., & Schlozman, K. L. (1995). Beyond ses: a resource model of political participation. The American Political Science Review, 89(2), 271–294.
http://doi.org/10.2307/2082425
Campbell, A., Gurin, G., & Miller, W. E. (1954). The voter decides. Evanston, Illinois: Row, Peterson and Company. http://doi.org/doi: 0.1177/000271625429600134
Craig, S. C., Niemi, R. G., & Silver, G. E. (1990). Political efficacy and trust: a report on the NES pilot study items. Political Behavior, 12(3), 289–314.
De Swert, K., Schacht, L., & Massini, A. (n.d.). More than a human tragedy? Italian Studies. Deutsche Presse Agentur. (2015, January 27). Zunahme der gewalt gegen migranten seit beginn
von pegida. Increase of violence towards migrants since pegida. Retrieved June 16, 2015, from http://www.wiwo.de/politik/deutschland/rechtsextremismus-zunahme-der-gewalt-gegen-migranten-seit-beginn-von-pegida-/11290626.html
Donsbach, W. (2014, December 18). „Wort im mund umdrehen“. ‘To twist words’. Retrieved June 17, 2015, from http://www.tagesspiegel.de/medien/pegida-und-die-luegenpresse-wort-im-mund-umdrehen/11140250.html
Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Orlando, FL: Harcourt Brace Jovanovich College Publishers.
Feldman, L. (2014). The hostile media effect. In K. Kenski & K. H. Jamieson, Oxford Handbook of Political Communication. http://doi.org/DOI: 10.1093/oxfordhb/9780199793471.013.011
Feldman, L., Hart, P. S., Leiserowitz, A., Maibach, E., & Roser-Renouf, C. (2015). Do hostile media perceptions lead to action? Communication Research. http://doi.org/DOI:
10.1177/0093650214565914
Field, A. (2013). Discovering statistics using IBM SPSS statistics (2nd ed.). London: Sage. Fishkin, J. S. (1997). The voice of the people: public opinion and democracy. New Haven: Yale
University Press.
Gunther, A. C., & Liebhart, J. L. (2006). Broad reach or biased source? decomposing the hostile media effect. Journal of Communication, 56(3), 449–466. http://doi.org/10.1111/j.1460-2466.2006.00295.x
Gunther, A. C., & Schmitt, K. (2004). Mapping boundaries of the hostile media effect. Journal of Communication, 54(1), 55–70. http://doi.org/10.1111/j.1460-2466.2004.tb02613.x
Gunther, A. C., Christen, C. T., Liebhart, J. L., & Chia, S. C.-Y. (2001). Congenial public, contrary press, and biased estimates of the climate of opinion. Public Opinion Quarterly, 65(3), 295–320. http://doi.org/10.1086/322846
Hansen, G. J., & Kim, H. (2011). Is the media biased against me? A meta-analysis of the hostile media effect research. Communication Research Reports, 28(2), 169–179.
http://doi.org/10.1080/08824096.2011.565280
Hayes, A. F. (2012). PROCESS: A versatile computational tool for observed variable mediation, moderation and conditional process modeling. Retrieved June 6, 2015, from
http://www.afhayes.com/public/process2012.pdf
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis. New York: Guilford Press.
Hessling, R. M., Traxel, N. M., & Schmidt, T. J. (2004). Ceiling Effect. (M. S. Lewis-Beck, A. Bryman, & T. F. Liao). Thousand Oaks, CA: Sage Publications.
http://doi.org/10.4135/9781412950589.n102
Ho, S. S., Binder, A. R., Becker, A. B., Moy, P., Scheufele, D. A., Brossard, D., & Gunther, A. C. (2011). The role of perceptions of media bias in general and issue-specific political
participation. Mass Communication and Society, 14(3), 343–374. http://doi.org/10.1080/15205436.2010.491933
Hwang, H., Pan, Z., & Sun, Y. (2008). Influence of hostile media perception on willingness to engage in discursive activities: An examination of mediating role of media indignation. Media Psychology, 11(1), 76–97. http://doi.org/10.1080/15213260701813454
Imai, K., Tingley, D., & Yamamoto, T. (2013). Experimental designs for identifying causal mechanisms. Journal of the Royal Statistical Society: Series a (Statistics in Society), 176(1),
5–51.
Jiménez, M., Aguilar, R., & Alvero-Cruz, J. R. (2012). Effects of victory and defeat on testosterone and cortisol response to competition: Evidence for same response patterns in men and women. Psychoneuroendocrinology, 37(9), 1577–1581.
http://doi.org/10.1016/j.psyneuen.2012.02.011
Kaplan, J. T., Freedman, J., & Iacoboni, M. (2007). Us versus them: Political attitudes and party affiliation influence neural response to faces of presidential candidates. Neuropsychologia, 45(1), 55–64. http://doi.org/10.1016/j.neuropsychologia.2006.04.024
Kivlighan, K. T., Granger, D. A., & Booth, A. (2005). Gender differences in testosterone and cortisol response to competition. Psychoneuroendocrinology, 30(1), 58–71.
http://doi.org/10.1016/j.psyneuen.2004.05.009
Knutson, K. M., Wood, J. N., Spampinato, M. V., & Grafman, J. (2006). Politics on the brain: An fMRI investigation. Social Neuroscience, 1(1), 25–40.
http://doi.org/10.1080/17470910600670603
LaPiere, R. T. (1934). Attitudes vs. actions. Social Forces, 13(2), 230–237.
Lazarus, R. S. (1991). Progress on a cognitive-motivational-relational theory of emotion. American Psychologist, 46(8), 819.
Lobo, S. (2015, July 8). Griechenland, flüchtlinge, AfD: Deutschlands neo-nationalismus. Greece, refugees, AfG: Germany's neo-nationalism. Retrieved July 9, 2015, from http://www.spiegel.de/netzwelt/web/deutschlands-neo-nationalismus-lobo-kolumne-a-1042600.html
Maio, G. R., Olson, J. M., Bernard, M. M., & Luke, M. A. (2006). Ideologies, values, attitudes, and behavior. In Handbooks of Sociology and Social Research (pp. 283–308). Springer US.
http://doi.org/10.1007/0-387-36921-X_12
Matheson, K., & Dursun, S. (2001). Social identity precursors to the hostile media phenomenon: partisan perceptions of coverage of the bosnian conflict. Group Processes & Intergroup Relations, 4(2), 116–125. http://doi.org/10.1177/1368430201004002003
Matthes, J., Morrison, K. R., & Schemer, C. (2010). A spiral of silence for some: Attitude certainty and the expression of political minority opinions. Communication Research, 37(6), 774–800. http://doi.org/10.1177/0093650210362685
Mehta, P. H., & Josephs, R. A. (2006). Testosterone change after losing predicts the decision to compete again. Hormones and Behavior, 50(5), 684–692.
http://doi.org/10.1016/j.yhbeh.2006.07.001
Moskalenko, S., & McCauley, C. (2009). Measuring political mobilization: The distinction between activism and radicalism. Terrorism and Political Violence, 21(2), 239–260. http://doi.org/10.1080/09546550902765508
Mudde, C. (2015, March 2). Local shocks: the far right in the 2014 European elections. Access Europe. Amsterdam.
Neiman, J., Smith, K. B., French, J., Waismel-Manor, I., Israel, & Hibbing, J. R. (2013). Is voting at the polls more stressful than voting at home? A field experiment. Presented at the APSA Annual Meeting.
Newhagen, J. E. (1998). TV news images that induce anger, fear, and disgust: Effects on approach‐ avoidance and memory. Journal of Broadcasting & Electronic Media, 42(2), 265–276. http://doi.org/10.1080/08838159809364448
Niemi, R. G., Craig, S. C., & Mattei, F. (1991). Measuring internal political efficacy in the 1988 national election study. The American Political Science Review, 85(4), 1407–1413.
http://doi.org/10.2307/1963953
Perloff, R. M. (1989). Ego-involvement and the third person effect of televised news coverage. Communication Research, 16(2), 236–262.
Perloff, R. M. (2015). A three-decade retrospective on the hostile media effect. Mass Communication and Society, 1–42. http://doi.org/10.1080/15205436.2015.1051234 Rojas, H. (2010). “Corrective” Actions in the Public Sphere: How Perceptions of Media and
Media Effects Shape Political Behaviors. International Journal of Public Opinion Research, 22(3), 343–363. http://doi.org/10.1093/ijpor/edq018
Rule, N. O., Freeman, J. B., Moran, J. M., Gabrieli, J. D. E., Adams, R. B., & Ambady, N. (2010). Voting behavior is reflected in amygdala response across cultures. Social Cognitive and Affective Neuroscience, 5(2-3), 349–355. http://doi.org/10.1093/scan/nsp046
Ryan, T. J. (2012). What makes us click? Demonstrating incentives for angry discourse with digital-age field experiments. The Journal of Politics, 74(04), 1138–1152.
http://doi.org/10.1017/S0022381612000540
Scheufele, D. A., & Moy, P. (2000). Twenty-five years of the spiral of silence: a conceptual revieaw and empirical outlook. International Journal of Public Opinion Research, 12(1), 3– 28. http://doi.org/10.1093/ijpor/12.1.3
Schulz, A. (2013). All hostile media. (M. Wettstein).
Statistisches Bundesamt. (2013). Bildungsstand. Retrieved June 16, 2015, from
https://www.destatis.de/DE/ZahlenFakten/GesellschaftStaat/BildungForschungKultur/Bildun gsstand/BildungsstandInfo.html
Statistisches Bundesamt. (2014, April 10). Bevölkerungsstand. Retrieved June 17, 2015, from https://www.destatis.de/DE/ZahlenFakten/GesellschaftStaat/Bevoelkerung/Bevoelkerungssta
nd/Tabellen/Zensus_Geschlecht_Staatsangehoerigkeit.html
SZ.de/cag. (2015, January 13). "Lügenpresse" ist unwort des jahres 2014. “Lying press” is non-word of the year 2014. Retrieved May 17, 2015, from
http://www.sueddeutsche.de/kultur/sprache-luegenpresse-ist-unwort-des-jahres-1.2295042 Theiss-Morse, E., & Hibbing, J. R. (2005). Citizenship and civic engagement. Annu Rev Polit Sci,
2005(8), 227–249. http://doi.org/10.1146/annurev.polisci.8.082103.104829
Thompson, D. F. (2008). Deliberative democratic theory and empirical political science. Annu Rev Polit Sci, 11(1), 497–520. http://doi.org/10.1146/annurev.polisci.11.081306.070555
Tsfati, Y., & Cohen, J. (2005). Democratic consequences of hostile media perceptions: The case of gaza settlers. The Harvard International Journal of Press/Politics, 10(4), 28–51.
Valentino, N. A., Brader, T., Groenendyk, E. W., Gregorowicz, K., & Hutchings, V. L. (2011). Election night’s alright for fighting: the role of emotions in political participation. The Journal of Politics, 73(01), 156–170. http://doi.org/10.1017/S0022381610000939
Valentino, N. A., Gregorowicz, K., & Groenendyk, E. W. (2008). Efficacy, emotions and the habit of participation. Political Behavior, 31(3), 307–330. http://doi.org/10.1007/s11109-008-9076-7
Valkenburg, P. M., & Peter, J. (2013). Five challenges for the future of media-effects research. International Journal of Communication, 7, 19.
Vallone, R. P., Ross, L., & Lepper, M. R. (1985). The hostile media phenomenon: Biased perception and perceptions of media bias in coverage of the beirut massacre. Journal of Personality and Social Psychology, 49(3), 577–585.
Verba, S., & Nie, N. H. (1987). Participation in america: political democracy and social equality. Chicago: University of Chicago Press.
Yap, A. J., & Tong, E. M. W. (2009). The appraisal rebound effect: Cognitive appraisals on the rebound. Personality and Social Psychology Bulletin, 35(9), 1208–1219.
Zinck, A., & Newen, A. (2008). Classifying emotion: A developmental account. Synthese, 161(1), 1–25.
Appendix A Appendix A1: Questionnaire
Factsheet
Amsterdam, 22. Mai, 2015 Lieber Teilnehmer,
Du bist herzlich eingeladen zur Teilnahme an einem Forschungsprojekt des ASCoR Instituts an der Universität Amsterdam.
Der Name des Forschungsprojektes lautet „Jährliches Update zu Medien und Politik in Europa“. Jede in Deutschland lebende Person über 16 kann an dieser Studie teilnehmen. Ziel der Studie ist es,
kontinuierlich - über mehrere Jahre hinweg - die Qualität von Journalismus und Politik zu messen, sowie die Reaktion der Bürger hierauf.
Im Zuge dieser Studie wirst Du an einer kurzen Umfrage teilnehmen. Das dauert ungefähr XX Minuten.
Da das Projekt unter der Verantwortlichkeit des ASCoR Institutes, Universität Amsterdam, steht, garantieren wir folgende Punkte:
1. Deine Anonymität ist gesichert und Deine Antworten und Daten werden unter keinen Umständen an Dritte weitergegeben
2. Du kannst Deine Teilnahme an der Studie jederzeit abbrechen. Zudem hast Du die Möglichkeit, 24 Stunden nach Teilnahme Deine Antworten zurückzuziehen.
3. Durch Deine Teilnahme wirst Du keinen Risiken oder unangenehmen Inhalten ausgesetzt 4. Spätestens fünf Monate nach Abschluss des Projektes können wir Dir einen Bericht über die Ergebnisse der Studie zur Verfügung stellen
Wir hoffen, wir haben Dich hinreichend informiert. Wir bedanken uns im Voraus für Deine Hilfe in diesem Projekt. Deine Mitarbeit ist für uns sehr wertvoll.
Mit besten Grüßen, Laura Schacht
Informed Consent
Hiermit erkläre ich, dass ich unmissverständlich über die Art und Methode des Forschungsprojekts, wie auf der vorherigen Seite beschrieben, informiert wurde.
Ich stimme vollkommen und freiwillig zu, an dieser Studie teilzunehmen. Mir bleibt jedoch das Recht vorbehalten, meine Zustimmung ohne Angabe von Gründen zurückzuziehen. Auch kann ich die Teilnahme an der Umfrage zu jeder Zeit abbrechen. Sollten die Ergebnisse dieser Studie
wissenschaftlich oder auf eine andere Art und Weise publiziert werden, so wird die Anonymität der Teilnehmer sichergestellt ist. Darüber hinaus werden meine persönlichen Daten nicht ohne mein ausdrückliches Einverständnis an Dritte weitergegeben.
Für den Fall, dass ich weitere Informationen zu diesem Forschungsprojekt erhalten möchte, kann ich mich jederzeit an die Projektleiterin Laura Schacht (laura.schacht@student.uva.nl, ASCoR, University of Amsterdam, Nieuwe Achtergracht 166, 1018 WV Amsterdam) wenden.
Bzgl. Beschwerden, Anmerkungen o.ä. haben, kann ich mich an das Ethikkommittee der Amsterdam School of Communication Research (ASCoR) der Universität Amsterdam unter folgender Adresse wenden: ASCoR secretariat, Ethics Committee, University of Amsterdam, Roeterseilandcampus, Nieuwe Achtergracht 166, 1018 WV Amsterdam; 020-5253680; ascor-secr-fmg@uva.nl.
Coverstory
Lieber Teilnehmer
Jedes Jahr erhebt die Universität Amsterdam zusammen mit 7 Partneruniversitäten in ganz Europa eine Studie über sozial relevante Themen in den Medien und wie die Politik darauf reagiert. Dies ist wichtig um der Gesellschaft und den Bürgern Kontrollmöglichkeiten zu bieten.
In der folgenden Umfrage werden Sie einen Artikel aus 5 verschiedenen Themenbereichen lesen. Diese Artikel repräsentieren die Mehrheit der Artikel, die in deutschen Medien zu diesem Thema zu finden sind. Wir bitten Sie, diesen Artikel aufmerksam und komplett zu lesen, damit Sie im Anschluss Fragen dazu beantworten können. Folgend würden wir Ihnen gerne einige Fragen über Politik stellen, bevor wir etwas mehr über Sie im letzten Teil erfahren möchten.
Im Voraus schon einen herzlichen Dank für Ihre Hilfe und Zeit. Laura Schacht
A: Controls
Media Use
Variable Name: A2_type
Question: Was sind die Nachrichtenmedien die Sie am häufigsten Nutzen, um Nachrichten zu sehen / lesen? Sie können mehrere Optionen anwählen:
Items:
(1) Öffentlich-Rechtliches Fernsehen (2) Privates Fernsehen
(3) Internet
(4) Individuell generierte Emails
(5) Radio
(6) Tageszeitungen (7) Wochenzeitungen
(8) Magazine
(9) Andere: [STRING] Variable Name: A3_frequency
Question: Wie viele Stunden pro Woche konsumieren sie im Durchschnitt Nachrichten?
Coding: ( 1) Nie. Ich schaue/lese keine Nachrichten
(2) Falls irgendwo eine Zeitung herumliegt, schaue ich vielleicht mal rein, aber das ist seltener als einmal in der Woche
(3) Weniger als eine Stunde pro Woche (4) Eine oder zwei Stunden pro Woche (5) Zwei bis drei Stunden pro Woche (6) Vier bis sieben Stunden pro Woche
(7) Mehr als sieben Stunden pro Woche. Also im Durchschnitt mehr als eine Stunde pro Tag.
State
Variable Name: A4_state
Question: In welchem Bundesland leben Sie zurzeit?
Coding: (1) Baden-Württemberg
(3) Berlin (4) Brandenburg (5) Bremen (6) Hamburg (7) Hessen (8) Mecklenburg-Vorpommern (9) Niedersachsen (10) Nordrhein-Westfalen (11) Rheinland-Pfalz (12) Saarland (13) Sachsen (14) Sachsen-Anhalt (15) Schleswig-Holstein (16) Thüringen B: Political Efficacy
Political Efficacy Scale
Variable Name: B1_poleffi
Question: Nun geht es um Ihre Einschätzung zu politischen Fragen. Den folgenden Aussagen können Sie mehr oder weniger zustimmen. Inwieweit stimmen Sie der jeweiligen Aussage zu?
Items: (1) Wichtige politische Fragen kann ich gut verstehen und einschätzen. (3) Ich traue mir zu, mich an einem Gespräch über politische Fragen aktiv zu
beteiligen.
Coding: (1) stimme gar nicht zu (2) stimme nicht zu (3) stimme eher nicht zu
(4) stimme weder zu noch dagegen (5) stimme eher zu
(6) stimme zu
Stimuli Material
Vorausgegangene Forschungen haben ergeben, dass der folgende Nachrichtenbeitrag der Mehrheit von Beiträgen entspricht, wie sie in deutschen Medien in den letzten 6 Monaten publiziert wurden. Das Thema dieses Beitrages ist: Immigration und Flüchtlinge.
Im weiteren Verlauf der Umfrage werden wir Ihnen Fragen zu diesem Artikel stellen. Bitte lesen Sie daher den Artikel aufmerksam und vollständig durch.
[ARTICLE]
C: Hostile Media Perception
Participants Ideology General
Variable Name: C1_ideology
Question: Zunächst eine Frage an Sie persönlich, im Anschluss folgen Fragen zu Medien allgemein und dem Artikel den Sie zuvor gelesen haben:
Bitte geben Sie auf der unten stehenden Skala an wo Sie sich selber im politischen Spektrum sehen.
Coding: (1) sehr links / liberal (2-6) mitte
(7) sehr rechts / konservativ HMP General
Variable Name: C2_hmpgeneral
Question: Nun würden wir gerne von Ihnen wissen, wie Sie die politische Positionierung der Medien in ihrer Berichterstattung allgemein einschätzen.
Coding: (1) sehr links / liberal (2-6) mitte
(7) sehr rechts / konservativ HMP Specific
Variable Name: C3_hmpspec
Question: Die folgenden Fragen beziehen sich auf den Artikel, den Sie am Anfang gelesen haben. Würden Sie sagen, dass der Artikel eher ausgewogen über Immigration und Flüchtlinge berichtet, oder dass er unausgewogen in die eine oder in die andere Richtung
Coding: (1) sehr parteiisch gegen Immigration und Flüchtlinge (2-6) neutral
(7) sehr parteiisch für Immigration und Flüchtlinge
Variable Name: C4_hmpspec
Question: Würden Sie sagen, dass der Reporter, der den Artikel verfasst hat, neutral ist in Bezug auf Immigration und Flüchtlinge, oder dass er parteiisch ist?
Coding: (1) sehr parteiisch gegen Immigration und Flüchtlinge (2-6) neutral
(7) sehr parteiisch für Immigration und Flüchtlinge
Variable Name: C5_hmpspec
Question: Bitte versuchen Sie, sich an den Artikel zu erinnern. Zu wie viel Prozent war der Artikel, den Sie gelesen haben, für Immigration und Flüchtlinge und zu wie viel Prozent
dagegen?
Items. (1) % Vorteilhaft für Immigration und Flüchtlinge (2) % Nachteilig für Immigration und Flüchtlinge
D: Emotion
Anger / Anxiety / Schadenfreude / Disgust
Variable Name: D1_emotion
Question: Das Konsumieren von Nachrichten ruft in Menschen verschiedene Emotionen hervor. Bitte geben Sie für die unten gelisteten Emotionen an, in wie weit Sie diese verspürt haben, während oder nach dem Lesen des Artikels:
Items: (1) Wut
(2) Ekel
Coding: (1) stimme gar nicht zu (2) stimme nicht zu (3) stimme eher nicht zu
(4) stimme weder zu noch dagegen (5) stimme eher zu
(6) stimme zu
Participation
In dem folgenden Teil des Fragebogens würden wir gerne mehr darüber wissen, wie Sie auf Politik reagieren.
E: Self Evaluation Measures of Intended Participation
Personal Evaluation of Actions (Offline, General)
Variable Name: E1_actioneval
Question: In einer Demokratie wie Deutschland gibt es verschiedene Wege Dinge zu verbessern und die Gesellschaft davor zu bewahren, dass manche Dinge falsch laufen.
Wie wahrscheinlich wäre es für Sie persönlich, dass Sie sich an den folgenden Aktivitäten beteiligen würden, wenn es um Immigration und Flüchtlinge geht?
Items: (1) Ich würde an einer Demonstration/einem öffentlichen Protest teilnehmen. (2) Ich würde die Anfahrt von einer Stunde in Kauf nehmen, um an einer
Demonstration/einem öffentlichen Protest teilzunehmen.
(3) Ich würde an einer Demonstration/einem öffentlichen Protest teilnehmen. Auch wenn die Demonstration eventuell gewaltsam verläuft.
(4) Ich würde Polizei, Sicherheitsleute oder Gegendemonstranten angreifen, sollte ich sehen wie diese Demonstranten schlagen, die meine Ansichten vertreten. (5) Ich würde an öffentlichen Diskussionen oder Vorlesungen zum Thema
teilnehmen.
(6) Ich würde mich zu ehrenamtlicher Arbeit bereit erklären (z.B. Petitionen
schreiben, Flyer und Informationsblätter verteilen, andere Freiwillige rekrutieren, etc.)
(7) Ich würde einer Organisation, die meine Ansichten und Werte vertritt, Geld spenden.
(8) Ich würde die Organisation, die meine Ansichten und Werte vertritt, auch weiterhin unterstützen. Selbst wenn diese Organisation manchmal die Gesetze bricht und manchmal Gewalt anwendet.
Coding: (1) sehr unwahrscheinlich (2-6)
F: Direct Participation Online
Facebook Page Membership (Online; IS)
Variable Name: F1_fbpage
Question: Stellen Sie sich vor, Sie loggen sich jetzt bei Facebook ein und die folgenden Seiten werden Ihnen vorgeschlagen. Haben Sie Interesse daran, einer oder mehrerer dieser Seiten zu folgen?
Wenn sie kein Facebook Profil haben; geben Sie an ob sie Interesse daran hätten von diesen Seiten Informationen abzurufen.
Options: (1)
(2)
(3)
(5)
(6)
(7)
Coding: (1) ja
Petition (Online; IS)
Variable Name: F4_petition
Question: Petitionen sind eine Möglichkeit, um Ihrer Stimme in der Politik Gehör zu verschaffen und Ihre Stimme in der Öffentlichkeit zu erheben. Wollen Sie eine der folgenden Petitionen (oder mehrere) nach dem Beenden dieses Fragebogens unterschreiben?
Options: (1)
(2)
(3)
(4)
