The threat of audience fragmentation as a consequence of an increasingly diverse media landscape has long been a subject of discussion. In the internet age, the debate has intensified further due to the explosion in the range of content on offer and the changes in how content is presented and disseminated, both in terms of content and technology. From the user’s perspective, online platforms offer far greater opportunities for selective – and therefore more interest-driven – news consumption. From the perspective of democratic theory, the consequences of individualised consumption are predominantly interpreted as negative: it is assumed that audience fragmentation increases the risk of disintegration. Alongside the conscious selection made by media users, unnoticed, technologically driven pre-selection now also plays an important and, as yet, largely unexplored role. Information intermediaries such as search engines, news aggregators and social networks act as intermediaries between content providers and users, unconsciously guiding the latter in their choice of news. This is because they collect, structure, weight and aggregate information, thereby controlling the degree to which topics can be found. Whilst they provide users with welcome guidance and navigation assistance, they also harbour new potential for influence, arising from algorithm-based weighting mechanisms such as the personalisation of search results. Although the current debate on the social role of intermediaries is conducted almost exclusively in critical terms, there is a lack of clear evidence regarding both negative and positive effects.
It therefore remains to be seen to what extent automated selection mechanisms exacerbate or mitigate tendencies towards fragmentation. This DFG project aims to fill this research gap. The aim of the study is to determine the influence of the intermediaries’ weighting criteria, both individually and in combination, on the degree of audience fragmentation. The theoretical contribution lies in a network-theoretical modelling of individual news selection across several levels of analysis, which highlights the influence of the intermediaries. The empirical core of the project consists of an innovative combination of methods: content analysis is used to survey the range of journalistic topics covered by the leading German online news outlets, whilst representative tracking data is used to examine the extent to which users are actually exposed to these topics. Through this combination of methods, the study provides, for the first time, a picture of the diversity and fragmentation of content provision and usage, and is able to offer a realistic assessment of the much-discussed ‘filter bubble’. In doing so, it contributes to measuring the wider societal impacts of algorithm-based information consumption via search engines or social networks.
Project manager:
Project team members:
Funding:
German Research Foundation (STA 1437/3-1)
Duration: 2017–2021
