If Network Heterogeneity Is Important for Information Diets, What Are Its Causes?

The second presentation in this IAMCR 2019 session is presented by Nadine Strauß, whose focus is on the approaches by news readers to exposing themselves to a diversity of viewpoints. To do so is important for democracy, but it seems that polarisation in society is increasing, and there remain concerns about the role of ‘filter bubbles’ on people’s information diets.

But political beliefs, attitudes, and even voting behaviours still remain strongly influenced by people’s personal and familial networks rather than just by their online and social media activities; here, network heterogeneity plays a critical role in ensuring the diversity of people’s information diets.

What factors might be able to predict just how heterogeneous a person’s network will be? This will rely in part on demographics; on the structure of their networks; on political behaviours, interests, perceptions of self-efficacy, and strength of ideological beliefs; and on news media use practices. Finally, the frequency of political discussions with others might also mediate the effects of such attributes.

The present project builds on a demographically representative two-wave online survey in the U.S., in 2013-14. It found that age matters: older people have less heterogenous social networks; social media users and people with larger networks encounter more heterogeneous content; offline participation, political interest and knowledge, and political efficacy also have some small effects; use of traditional news helps a little as well; and discussion frequency is especially important as a mediating factor. The more people talk about politics and the more knowledgeable they are about it, the less they actively seek out heterogeneous networks, however.

One suggestion from this work is to create more platforms in society – across many fields of social engagement – that enable talk about politics and public affairs. There is also a question of how these patterns develop over time, of course – the present data are from 2013 and 2014, and U.S. politics has moved on considerably since then. More qualitative studies may also help to flesh out this picture further.