Allgemein

Fragmentation of the audience due to information intermediaries

Audience fragmentation has long been feared to result from an increasing fragmentation of content supply. In the Internet age, the debate has become even more heated due to the explosion in content and the changes in the way content is presented and distributed. From the user's point of view, online media provide far more opportunities to practice active selection and thus more interest-based news consumption. In terms of democratic theory, the hypothesized consequences of individualized news consumption are predominantly negative: A fragmentation of audiences is believed to increase the risk of societal disintegration. In addition to active selection by media users, subtle and technically controlled pre-selection now also plays an important and hitherto largely unexplored role. Information intermediaries such as search engines, news aggregators and social networks act as brokers between the media content supply and the user, unconsciously guiding the latter in her choice of news. By collecting, structuring, weighting and aggregating  content, they control the degree to which issues are encountered. For the users, they act as welcome orientation and navigation aids. At the same time, they influence these users through algorithm-based weighting logics such as the personalization of search results. Although the current debate on the societal role of information intermediaries is almost exclusively critical, there is a scarcity of clear evidence of both negative and positive effects.

The extent to which automated selection mechanisms reinforce or mitigate fragmentation tendencies is therefore an open question. The project aims to close this research gap. The goal of the study is therefore to determine the influence of the different weighting logics of the intermediaries, individually and in combination, on the degree of fragmentation of the audience. In theoretical terms, the project is based on a network-theoretical modelling of individual content selection on several levels of analysis, which makes the influence of the intermediaries visible. The empirical core of the project is an innovative combination of methods: By means of a content analysis, the journalistic range of topics of the most important German online news media is analyzed, and by means of representative tracking data, the extent to which users are actually confronted with these topics is investigated. This combination of methods enables the study to provide a picture of the diversity of supply and exposure diversity or fragmentation for the first time and to make a realistic assessment of the much-discussed filter bubble. It thus contributes to measuring the effects of algorithm-based news consumption via search engines or social networks on society as a whole.

Head of project:
Prof. Dr. Birgit Stark

Project staff:
Dr. Pascal Jürgens

Funding:
German Research Foundation (STA 1437/3-1)
Period: 2017-2021

Posted on | Posted in Allgemein

Social Bots and Fake News in the German federal elections 2017

The Ukrainian crisis, the Brexit referendum and the US election campaign 2016 have shown that political decision-making on the Internet is facing new threats: In particular, so-called social bots use machine automation to autonomously search for topics in social networks, spread articles and make comments that are difficult to distinguish from those of real users. There is a credible danger that social bots can generate attention for extreme positions or trigger a shift in public discourse through the sheer mass of automated content supply and interactions.

But to what extent do computer programs disguised as humans actually manipulate political discourse on Facebook? Can their mass-produced contributions to discussions influence the public formation of opinion and even political decision-making processes? And what role does the phenomenon of "fake news" play in the context of these algorithm-based communication processes? The research project is dedicated to these questions within the context of the German federal election of 2017.

With regard to potential manipulative interventions, three independent effects are examined: (1) the generation of attention, agenda setting and the shaping of images as a direct influence on users through news items (re-)produced by bots, (2) the influence on the perceived climate of opinion by strengthening certain viewpoints, and (3) the influence on metrics (e.g., number of comments), which often serve users as orientation points for the perception and evaluation of discussions.

In addition, the phenomenon of "fake news" will be examined more closely. These messages, predominantly with an affective load and strongly biased reporting, exploit  the interaction-oriented mechanisms of social media platforms, which thus may fuel conspiracy theories. On the basis of initial investigations, various effects of fake news on online discussions are also being examined: (1) the direct influence of users who believe the widespread fake news, (2) the polarization of the public through the publication and sharing of fake news on Facebook, and (3) the potential influence of emotional style and explosiveness of news on a more frequent and longer lasting further dissemination of these news and a resulting higher probability of accidental exposure by citizens who are not critical of elites and the media.

head of project:
Prof. Dr. Marcus Maurer
Prof. Dr. Christian Schemer
Prof. Dr. Birgit Stark

Responsible research staff:
Dr. Pascal Jürgens
Simon Kruschinski M.A.

Funding:
research program media convergence

Posted on | Posted in Allgemein