The filter bubbles are not as hermetic as one may think

The influx of refugees in 2015 put social cohesion in Austria to the test. With funding from the Austrian Science Fund FWF, the sociologist Jörg Flecker from the University of Vienna and his team conducted a representative survey to explore how ideas of solidarity are negotiated in times of crisis. The survey confronted a representative sample of individuals with tough questions: is solidarity important or controversial? Who is entitled to benefits from the welfare state? What people are entitled to support and to what extent? And what are the requirements for such support?

In 2018, building on the basic-research project, principal investigator Flecker then collaborated with postdoctoral researcher Saskja Schindler and citizen scientists to investigate how social cement and the associated political orientation are negotiated in social media. “We wanted to find out how these issues are discussed when different opinions come together. Social media harbour the danger of creating filter bubbles and echo chambers in which one's own views are always reinforced. We wanted to take a closer look at that aspect,” explains Jörg Flecker. And Saskja Schindler adds: “The citizen science approach, i.e. working with lay researchers, is particularly useful here because the aim is to explore life worlds that are as diverse as possible, worlds to which one would hardly have access otherwise.”

The results are both methodologically interesting and also show that the echo chambers do not constantly throw back the same things from every wall, and that it may well be worth the effort to try and change people’s opinions in digital forums. The project was carried out in close cooperation with FORBA – Working Life Research Centre Vienna – and Veronika Wöhrer, a specialist in participatory research at the University of Vienna.

Start by settling questions of research ethics

In order to gain insights into different milieus, project coordinator Saskja Schindler accessed different Facebook groups: “We could not simply post our call to participate in the research into groups spanning the whole gamut from rifle guilds to lobbying groups and the anti-trust association. Quite often the admins did not react at all or react negatively. That was surprising, given that people often have quite open discussions in these forums.” Hence the team directly addressed frequent posters on the respective pages via direct messages in order to recruit ten citizen scientists who should differ in political orientation, gender, background, age, employment status, etc.

The second challenge was to ensure informed consent, i.e. making all participants aware at all times that they were part of a research project. This was achieved by providing information in the group pictures that were always visible. After a training session, the citizen scientists independently initiated discussions in closed groups with invited groups of friends and then made the written communication available in a carefully pseudonymised format. The topics chosen as being liable to provoke hot discussions were “an hourly wage of EUR 1.50 for asylum seekers doing community service” and “the new means-tested minimum income scheme”.

Conclusions for social media research

Ultimately, the researchers were able to recruit nine citizen scientists. The discussions showed great differences. A few did not really get going, for example because the groups of friends felt no connection to the topic. The methodological conclusions Jörg Flecker draws from the results are relevant for social media research: “Despite the promise of anonymity, the participants in the closed-group discussions felt less secure and were more hesitant in their statements than in the public discussions in the forums. One would expect this to be the other way round. Our explanation is that the regular feeds are perceived as more transient. Posts slide down the timeline and disappear from view, whereas in our format it was clear: someone is still reading this, using it for research. There was probably a perception of more pronounced social control.”

How airtight is the filter bubble?

One conclusion of the project is that it would be an unjustified generalisation to consider groups that exchange views in the social media to be nothing more than filter bubbles and echo chambers. The discussions held in the different circles of friends were relatively controversial and were not only exchanges of like-minded opinions. Even when people had similar ideas about why someone should receive support, opinions about how much support was appropriate were liable to diverge widely. The results were yet another confirmation that notions of justice play an important role in the approach to solidarity, i.e. whether someone becomes eligible for support and affiliation on the basis of performance, need or status. Even if the participants agreed that those who work should also reap benefits, there could be heated discussions about what exactly counts as work.

Potential for impacting opinions

Saskja Schindler emphasises that despite the targeted selection of participants, it is not the case that everyone always agrees. On the other hand, it also becomes apparent that social media are not a particularly suitable forum for general socio-political discussions: “Discussions remained at the level of concrete opinions on individual cases and facts. Things that need long and complex consideration tend to be ignored. Even if someone addressed higher-level concepts such as neoliberalism, there was usually no follow-up.” Also, it is simply tedious to read longer posts in the format provided by Facebook. The research team does see, however, that there is a potential for impacting opinions in regular digital get-togethers. You can enter into discussions and try to change other people's minds. And: circles of friends are not completely homogeneous on Facebook either, just as in real life, although the algorithm always promotes the greatest level of agreement.

Personal details

Jörg Flecker is Professor of General Sociology at the Institute of Sociology at the University of Vienna. His work focuses on work organisation and labour relations, the labour market, digitalisation, right-wing populism and extremism.

Saskja Schindler is a research associate at the SORA Institute for Social Research and Consulting and a university lecturer at the University of Vienna’s Department of Sociology. Her work focuses on the sociology of work, political sociology and social inequality.

The project Worlds apart? Solidarity concepts and political orientations in social media was supported by the Austrian Science Fund FWF with EUR 50,000 euros in the context of the Top Citizen Science funding initiative and ended in 2020. 

Publications

Wöhrer V., Schindler S., Papouschek U. & Schönauer A.: Protected by the Anonymous Mass? Reflecting Anonymity and Informed Consent in a Citizen Science Project on Social Media Platforms. Poster presented at the European Citizen Science Conference 2020 (PDF)

Hofmann J., Altreiter C., Flecker J., Schindler S., & Simsa R.: Symbolic struggles over solidarity in times of crisis: Trade unions, civil society actors and the political far right in Austria, in: European Societies, 21(5), 649–671, 2019

Altreiter C., Flecker J., Papouschek U., Schindler S., & Schönauer A.: Umkämpfte Solidaritäten. Spaltungslinien in der Gegenwartsgesellschaft, Promedia Verlag 2019