Final Report Summary - EMOCITI (Europe for the Hearts and Minds: The Role of the News Media and Emotions in Creating European Citizens)
The EMOCITI research project offers a novel understanding of how the news media influence European citizens’ political opinions and behaviour. Recent political communication research has shown that the news media’s potential to elicit emotions is one of the most important predictors of media effects. This means that even a short newspaper article that appeals to citizens’ anger about the EU is likely to affect people’s opinions and behaviours more than a longer article that uses factual information only. This effect is explained in the social psychological literature, which has shown that the presence of emotional language in a message, so-called ‘emotional cues’, functions as a catalyst of information processing and increases the persuasiveness of a message. The power of emotions is particularly important in the EU context. Studies have shown that the media frequently use emotional appeals in EU political news coverage – a characteristic that is only enhanced in today’s times of economic crisis. Yet, so far, only very few studies have experimentally shown which emotions have what specific effects on both individual opinions and behaviours. Also, no studies have tested how the role of emotions in the news media varies across different countries and political issues.
In this project, we made use of survey experiments to examine if and how emotion-eliciting news stories caused changes in citizens’ political perceptions and behaviour. The first experiment, conducted in the UK, tested how the presence of a range of positive of negative emotional cues in political news articles about elderly care influenced political opinions and behaviour. Based on a literature analysis, we made small adaptions to the experimental design in the EMOCITI project proposal. Because many political researchers are interested in the effects of in- and outgroups, we also tested if the type of actor mentioned within an emotion-eliciting news article (national, European or US) changed the extent to which participants reacted emotionally to this article. Additionally, we also included political behaviour (rather than just political opinions) as a dependent variable.
In a second experiment, we focused on emotions that had proven particularly interesting in the first experiment. Specifically, we tested if the use of emotions with varying temporal orientation (i.e. articles mentioning evaluative emotions such as relief and anger vs. articles using projective emotions such as hope and fear) plays a role in how news media influence opinions and behaviour. We conducted this second experiment in the US. This added a comparative aspect to the EMOCITI project and allowed us to evaluate if emotional responses are country-specific. In this second experiment, we tested emotional reactions to two different political issues in the news: elderly care and gun control. In this way, we were able to tap how different types of issues influence the role of emotions in the news.
Based on the results of the EMOCITI experiments, a third experiment will be conducted in autumn 2014 at the University of Amsterdam. In this third experiment, we will further develop our idea that there are different ‘types’ of emotional cues in the news.
Our findings suggest that positive emotions play a more powerful role in predicting media effects than previously expected. Most available studies focus on negative emotions such as anger and fear and have thus neglected the influence of emotions such as hope, pride and enthusiasm. The EMOCITI studies show that the use of positive emotional cues in the news not only causes stronger emotional reactions than the use of negative emotions, but can also be powerful mobilizer that will influence citizens’ political behaviour.
Our data also shows what role emotions play within the media effect process. We find that emotional reactions to a news article are mediators of effects on opinions and behaviours. Following the exposure to a news article, participants registered an emotional reaction which then led to changes in opinions and behaviours. Most media effect models only include cognitive mediators such as information processing or accessibility effects. The addition of emotional reactions as mediators allows for the development of a more comprehensive media effect model.
Furthermore, we find that, while effects on opinions are predominantly led by the valence of an emotional reaction, valence does not play the same role in effects on behaviours. If an article causes positive emotional reactions, then opinions become more positive (and vice versa when negative emotions are concerned). However, both positive and negative emotions can mobilize citizens to donate money to a cause, search for more information on the issue and even to vote in an election.
The data suggests that the effect processes we describe depend on both a person’s political predispositions (e.g. political cynicism or distrust) as well as character traits (e.g. how neurotic someone is). For instance, we find that political cynicism moderates emotional response in that those who are very cynical are ‘immune’ to the effects of positive emotional cues in the news media.
The data collected in this project is useful for the development of a typology of ‘emotional cues’ and their effects within the political news media. The studies conducted during the project period tested a wide range of emotions. Our survey experiments also included an extensive list of moderator variables. We also conducted our studies in different countries and will thus be able to add a comparative dimension to our research papers. We thus expect that our results will allow us to further explain how emotions function in the news media effects process. The role of emotions in political communication has attracted growing attention during recent years. Yet, there is so far only limited experimental data available.
Given the novel research designs applied in this project, our results are also valuable for the further development of the measurement of emotions and emotional traits in survey experiments. We measured a number of concepts from related fields (neuroscience, social psychology, political science) and are thus able to compare the validity and usefulness of these measures in the context of news media effects research.
Our data will contribute to the further theoretical integration of emotions into the field of political communication. However, results are also of great relevance to communication professionals, political actors and journalists. Our research may be used as a roadmap for the inclusion of emotional cues in political campaigns in the EU and beyond. Also, journalistic education and training may benefit from further scholarly insights as to how the use of emotional language during news production influences audiences. In the context of EU integration, our results may add to explanations of why citizens are (not) willing to support the European project now or in the future.
All results and further findings will be published in academic journals as well as on the project website (http://emociti.wordpress.com/) and the Research Fellow’s personal website: http://www.sophielecheler.com/.
In this project, we made use of survey experiments to examine if and how emotion-eliciting news stories caused changes in citizens’ political perceptions and behaviour. The first experiment, conducted in the UK, tested how the presence of a range of positive of negative emotional cues in political news articles about elderly care influenced political opinions and behaviour. Based on a literature analysis, we made small adaptions to the experimental design in the EMOCITI project proposal. Because many political researchers are interested in the effects of in- and outgroups, we also tested if the type of actor mentioned within an emotion-eliciting news article (national, European or US) changed the extent to which participants reacted emotionally to this article. Additionally, we also included political behaviour (rather than just political opinions) as a dependent variable.
In a second experiment, we focused on emotions that had proven particularly interesting in the first experiment. Specifically, we tested if the use of emotions with varying temporal orientation (i.e. articles mentioning evaluative emotions such as relief and anger vs. articles using projective emotions such as hope and fear) plays a role in how news media influence opinions and behaviour. We conducted this second experiment in the US. This added a comparative aspect to the EMOCITI project and allowed us to evaluate if emotional responses are country-specific. In this second experiment, we tested emotional reactions to two different political issues in the news: elderly care and gun control. In this way, we were able to tap how different types of issues influence the role of emotions in the news.
Based on the results of the EMOCITI experiments, a third experiment will be conducted in autumn 2014 at the University of Amsterdam. In this third experiment, we will further develop our idea that there are different ‘types’ of emotional cues in the news.
Our findings suggest that positive emotions play a more powerful role in predicting media effects than previously expected. Most available studies focus on negative emotions such as anger and fear and have thus neglected the influence of emotions such as hope, pride and enthusiasm. The EMOCITI studies show that the use of positive emotional cues in the news not only causes stronger emotional reactions than the use of negative emotions, but can also be powerful mobilizer that will influence citizens’ political behaviour.
Our data also shows what role emotions play within the media effect process. We find that emotional reactions to a news article are mediators of effects on opinions and behaviours. Following the exposure to a news article, participants registered an emotional reaction which then led to changes in opinions and behaviours. Most media effect models only include cognitive mediators such as information processing or accessibility effects. The addition of emotional reactions as mediators allows for the development of a more comprehensive media effect model.
Furthermore, we find that, while effects on opinions are predominantly led by the valence of an emotional reaction, valence does not play the same role in effects on behaviours. If an article causes positive emotional reactions, then opinions become more positive (and vice versa when negative emotions are concerned). However, both positive and negative emotions can mobilize citizens to donate money to a cause, search for more information on the issue and even to vote in an election.
The data suggests that the effect processes we describe depend on both a person’s political predispositions (e.g. political cynicism or distrust) as well as character traits (e.g. how neurotic someone is). For instance, we find that political cynicism moderates emotional response in that those who are very cynical are ‘immune’ to the effects of positive emotional cues in the news media.
The data collected in this project is useful for the development of a typology of ‘emotional cues’ and their effects within the political news media. The studies conducted during the project period tested a wide range of emotions. Our survey experiments also included an extensive list of moderator variables. We also conducted our studies in different countries and will thus be able to add a comparative dimension to our research papers. We thus expect that our results will allow us to further explain how emotions function in the news media effects process. The role of emotions in political communication has attracted growing attention during recent years. Yet, there is so far only limited experimental data available.
Given the novel research designs applied in this project, our results are also valuable for the further development of the measurement of emotions and emotional traits in survey experiments. We measured a number of concepts from related fields (neuroscience, social psychology, political science) and are thus able to compare the validity and usefulness of these measures in the context of news media effects research.
Our data will contribute to the further theoretical integration of emotions into the field of political communication. However, results are also of great relevance to communication professionals, political actors and journalists. Our research may be used as a roadmap for the inclusion of emotional cues in political campaigns in the EU and beyond. Also, journalistic education and training may benefit from further scholarly insights as to how the use of emotional language during news production influences audiences. In the context of EU integration, our results may add to explanations of why citizens are (not) willing to support the European project now or in the future.
All results and further findings will be published in academic journals as well as on the project website (http://emociti.wordpress.com/) and the Research Fellow’s personal website: http://www.sophielecheler.com/.
