Public Debate and Public Opinion during the Refugee Crisis in Germany
Wissenschaftszentrum Berlin, Germany
In 2015 Angela Merkel, a conservative head of state, made the highly liberal and unprecedented decision to open Germany’s borders for refugees while other EU countries kept them close. This decision ultimately led to the arrival of almost 1 million refugees in Germany in 2015 and caused heated public debate in print and film media. Traditionally, parties provide information heuristics for the public. People use parties to navigate complex political landscapes, and support policies and positions that their favored party puts forth. But what happens to public opinion, when traditional party lines blur in public debates, and politicians adopt views that are out of line with prior party decisions? In a survey experiment we presented 2 statements to a representative sample of German citizens (N=1209), one restrictive (“there is a massive abuse of the right to asylum”) and one liberal (“there is no upper right to the right to asylum”), and assessed their support for each of these statements. The statement had previously appeared in print and film media. We experimentally varied whether a party politician endorsed the statement or not (“…this statement [from a prominent SPD/CDU politician]”) thus creating three experimental groups for each statement: no endorsement, SPD endorsement, CDU endorsement. Our main finding was that participants agree the most with both statements - restrictive and liberal -in the control condition. They are particularly hesitant to support a statement if it comes from the party that is known for a position consistent with the statement, i.e., the SPD for the liberal statement and the CDU for the restrictive statement. We discuss these results in light of theories about reactance and resistance to polarization when people receive information from the media
Hostile Media Effect 2.0 - Influences of social networking sites on the (hostile) perception of news articles
University of Duisburg-Essen, Germany
Increasingly, news content is included by media organizations into social networking sites (SNS; Glynn, Huge, & Hoffman, 2012). An increase of incidental exposure to news and political content is more likely (Hasell & Weeks, 2016). According to PewResearchCenter (Gottfried, 2014) 86% of Facebook users see political content and news during social media use. People consume media daily but often perceive the media as biased and hostile to their own views (Oh, Park, & Wanta, 2011) in spite of actual even-handed news coverage. This phenomenon is called hostile media effect (HME) and was first documented by Vallone, Ross, and Lepper (1985). To examine the HME numerous studies have focused on newspapers, television or radio (e.g., Arpan & Raney, 2003; Choi, Yang, & Chang, 2009), blogs or Internet news sites (e.g., Matthes, 2011). However, to the best of our knowledge no research focuses on the HME within SNS. Therefore, this study focuses on the comparison of the HME between online news and news on a SNS. It was examined whether or not the environment of Facebook strengthened or weakened the perceived hostility in contrast to online news. An online experiment was conducted using a between-subjects design with four conditions varying the source of the news article (condition 1: news article was displayed as a Facebook post written by a close friend; condition 2: close friend shared news article of news service; condition 3: news service post; condition 4: news article on homepage). In order to be able to let the posts appear as being shared/posted by a close friend, participants were asked to name several Facebook friends (close friend, distant friend, etc.). The named close friend was then chosen to be displayed on the stimulus material. Tagesschau served as news service. The news article was identical for all conditions and the topic was the status of refugees in Germany. 212 people participated in this study. 28 participants were excluded because they stated a neutral opinion on the topic (see Choi et al., 2009) or made false statements regarding the source (n = 174, Mage= 27.5, SDage = 7.18, rangeage = 18 – 64, 107 female). In general, participants evaluated the news article as neutral (7-point likert scale, M = 4.56, SD = 1.03). Results of a ANOVA revealed no statistically significant differences between the conditions regarding participants’ perceived hostility of the news article (F(3,170) = .22, p = .885, ηp2 = .004). It seems that the SNS environment as well as the source do not have an influence on participants’ perceived article hostility. But, results of a ANOVA showed significant results between the conditions regarding different standards (correctness of arguments, fairness of arguments, relevance of information; F(3,170)= 3.12, p = .028, ηp2 = .052). A direct hostile media effect could not be found, but several explanatory mechanisms (e.g., prior beliefs) of the HME showed significant results.
“Piece by Piece or All in One Chunk?” How the Serialization of Scandalous Information Affects Recipients’ Attitudes Toward Political Candidates
U of Koblenz-Landau
Political scandals are on the rise and journalists tend to serialize political scandals publishing scandalous information bit by bit instead of all at once in a single news article. While the influences of repeated exposure to mediated stimuli have extensively been studied, the effects of serialization of political information have not been examined. Serialization(SE) and repetition of information(RE) can be conceptually differentiated. SE can be defined as a form of news coverage in which a journalist/media outlet withholds certain scandalous information and does not publish all available information at once but bit by bit. Thus, SE exposes recipients to a piece of information that is not presented as a whole, but piece by piece (e.g., several articles). In contrast, RE means that a piece of information is presented as a whole in a single presentation and, after some time, the same piece of information is presented again.
In a first step, (based on the respective literature) we developed a theoretical model on the effects of SE. According to the model, SE may serve as a subtle cue that elites/journalists assign importance to a certain issue. Thus, the greater the number of articles, the higher recipients’ perceived importance of the case (H1). Furthermore, it was expected that a higher degree of SE results in an increase in reading duration (H2) and (H3/H4) that both a higher perceived importance and an increase in reading duration will increase cognitive elaboration (e.g., thinking intensively about the depicted case). It was also assumed that the more intensively recipients elaborate on a political scandal, the more intensely they will feel negative emotions regarding the scandal (H5) and, finally (H6), the more intensely recipients feel negative emotions regarding a political scandal, the more negative their attitudes toward the political actor will be.
To test the hypotheses, we conducted an experiment (1x5 between-subject/N=171 students, M=23 years/SD=3.43/83% female) and exposed participants to a fictitious scandal (precluding any prior knowledge of the case). Participants were randomly assigned to one of five groups. The only difference between conditions was that the news was presented in 1/2/3/4 or 5 articles. The single-article group read a longer news article sub-divided in five sections, while, e.g., the five-article group read the identical information presented as five short articles captioned by the five sub-headings.
We tested the hypotheses by creating a structural equation model (AMOS23). Maximum likelihood estimation was used. The fit indices revealed good overall model fit. All path coefficients appeared to be statistically significant and pointing in the assumed directions (bias-corrected 95% confidence interval/5,000 bootstrap samples revealed an interval that was entirely below zero (β=-.02, CI[-.04,-.003]). That is, the serialization indeed indirectly negatively affected participants’ attitudes towards the politician via perceived importance, reading duration, cognitive elaboration, and the intensity of negative emotions. Interestingly, both mediating processes (perceived importance & reading duration) can be regarded as separate since both exerted independent influences on cognitive elaboration yet remained entirely uncorrelated (r=.01,ns). Implications are discussed.
The Personal is Political – The Influence of Politicians´ Facebook Status Updates on their Perception
1University Duisburg-Essen, Germany; 2University of Amsterdam
Nowadays social media channels like Facebook are not only used for private communication, but also as a source of news and political information (Bode, 2015). The lack of gatekeepers and the integration in user’s daily life are factors which strengthen the great potential of Facebook for politicians to get in touch with (potential) voters and to present themselves as ‘people like you and me’ by highlighting personal characteristics.
Research on social media in the political context is still in its infancy (Stieglitz et al., 2012) and especially the recipient’s perspective is understudied (Kruikemeier et al., 2013; Lee & Jang, 2011; Lee & Oh, 2012). An area of specific interest is how users perceive and evaluate politicians´ social media communication. First studies indicated that an interactive communication results in a more positive evaluation and higher voting intention (Kruikemeier, 2014; Lee & Jang, 2011; Thorson & Rodgers, 2006; Utz, 2009) and the combination of interactivity and personalization on a political website leads to feelings of reduced distance and greater connectivity (Kruikemeier et al., 2013).
The present study aims to extend the results by evaluating the impact of a personal communication strategy (in comparison to the publication of factual political information) used in politicians’ Facebook status updates on recipients’ perception. In this respect, we analysed the evaluation of politicians and the resulting parasocial interaction (PSI), both of which are assumed to be influenced by the personalization level of the status update and tested whether they have an impact on voting intention. In addition, by including a range of politicians, the study investigates whether there are differences in the effects depending on politicians’ gender, familiarity and party. 295 persons participated in the online survey and were randomly assigned to one of the politician’s Facebook pages with either a personal or a merely political status update.
Results showed an impact of the personalization level on the evaluation of the politician, in the way that a politician was evaluated as more trustworthy when he/she had posted a personal status update compared to a political one. Additionally, the reception of the personal posting led to a more intensive PSI, especially for recipients with a higher Facebook usage intensity. Moreover, the findings demonstrate mediation effects of trustworthiness and PSI on voting intentions: A personal status update led to a higher trustworthiness rating and PSI, which in turn increased voting intention.
Our study offers new insights regarding the integration of social media into political communication. Communicating with personal status updates offers more information about the politician and recipients are given a more complete impression of life, values and characteristics. This might be interpreted as reducing uncertainty leading to more trustworthiness, perceived intimacy and connectedness which in turn increases voting intention. Overall, the impact of personalization on the evaluation of the politician and PSI did not interact with the politicians’ gender, familiarity or party affiliation, which suggests that the positive effects of personal social media communication appears to be existing for a wide range of politicians.